HomeworkML.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "E1pF0okrM1I6"
      },
      "outputs": [],
      "source": [
        "import numpy as np # linear algebra\n",
        "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
        "\n",
        "url='https://drive.google.com/file/d/12dUeEO_rkButPD79pOteFoSSeuIBg5qt/view?usp=drive_link'\n",
        "file_id=url.split('/')[-2]\n",
        "dwn_url='https://drive.google.com/uc?id=' + file_id\n",
        "dataset = pd.read_csv(dwn_url)\n",
        "dataset = dataset.rename(columns={\"NObeyesdad\": \"NObesity\"})"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PN4MKJszM1I7"
      },
      "source": [
        "# About The Dataset\n",
        "The data consist of the estimation of obesity levels in people from the countries of Mexico, Peru and Colombia, with ages between 14 and 61 and diverse eating habits and physical condition , data was collected using a web platform with a survey where anonymous users answered each question, then the information was processed obtaining 17 attributes and 2111 records.\n",
        "The attributes related with eating habits are: Frequent consumption of high caloric food (**FAVC**), Frequency of consumption of vegetables (**FCVC**), Number of main meals (**NCP**), Consumption of food between meals (**CAEC**), Consumption of water daily (**CH20**), and Consumption of alcohol (**CALC**). The attributes related with the physical condition are: Calories consumption monitoring (**SCC**), Physical activity frequency (**FAF**), Time using technology devices (**TUE**), Transportation used (**MTRANS**)\n",
        "variables obtained :\n",
        "**Gender**, **Age**, **Height** and **Weight**.\n",
        "\n",
        "**NObesity** values are:\n",
        "\n",
        "* Underweight\n",
        "* Normal\n",
        "* Overweight\n",
        "* Obesity I\n",
        "* Obesity II\n",
        "* Obesity III"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "7vCgNeRzM1I8",
        "outputId": "1dc2cbf0-8255-4d47-ec15-e7710c674ab9"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      Gender        Age    Height      Weight family_history_with_overweight  \\\n",
              "0     Female  21.000000  1.620000   64.000000                            yes   \n",
              "1     Female  21.000000  1.520000   56.000000                            yes   \n",
              "2       Male  23.000000  1.800000   77.000000                            yes   \n",
              "3       Male  27.000000  1.800000   87.000000                             no   \n",
              "4       Male  22.000000  1.780000   89.800000                             no   \n",
              "...      ...        ...       ...         ...                            ...   \n",
              "2106  Female  20.976842  1.710730  131.408528                            yes   \n",
              "2107  Female  21.982942  1.748584  133.742943                            yes   \n",
              "2108  Female  22.524036  1.752206  133.689352                            yes   \n",
              "2109  Female  24.361936  1.739450  133.346641                            yes   \n",
              "2110  Female  23.664709  1.738836  133.472641                            yes   \n",
              "\n",
              "     FAVC  FCVC  NCP       CAEC SMOKE      CH2O  SCC       FAF       TUE  \\\n",
              "0      no   2.0  3.0  Sometimes    no  2.000000   no  0.000000  1.000000   \n",
              "1      no   3.0  3.0  Sometimes   yes  3.000000  yes  3.000000  0.000000   \n",
              "2      no   2.0  3.0  Sometimes    no  2.000000   no  2.000000  1.000000   \n",
              "3      no   3.0  3.0  Sometimes    no  2.000000   no  2.000000  0.000000   \n",
              "4      no   2.0  1.0  Sometimes    no  2.000000   no  0.000000  0.000000   \n",
              "...   ...   ...  ...        ...   ...       ...  ...       ...       ...   \n",
              "2106  yes   3.0  3.0  Sometimes    no  1.728139   no  1.676269  0.906247   \n",
              "2107  yes   3.0  3.0  Sometimes    no  2.005130   no  1.341390  0.599270   \n",
              "2108  yes   3.0  3.0  Sometimes    no  2.054193   no  1.414209  0.646288   \n",
              "2109  yes   3.0  3.0  Sometimes    no  2.852339   no  1.139107  0.586035   \n",
              "2110  yes   3.0  3.0  Sometimes    no  2.863513   no  1.026452  0.714137   \n",
              "\n",
              "            CALC                 MTRANS             NObesity  \n",
              "0             no  Public_Transportation        Normal_Weight  \n",
              "1      Sometimes  Public_Transportation        Normal_Weight  \n",
              "2     Frequently  Public_Transportation        Normal_Weight  \n",
              "3     Frequently                Walking   Overweight_Level_I  \n",
              "4      Sometimes  Public_Transportation  Overweight_Level_II  \n",
              "...          ...                    ...                  ...  \n",
              "2106   Sometimes  Public_Transportation     Obesity_Type_III  \n",
              "2107   Sometimes  Public_Transportation     Obesity_Type_III  \n",
              "2108   Sometimes  Public_Transportation     Obesity_Type_III  \n",
              "2109   Sometimes  Public_Transportation     Obesity_Type_III  \n",
              "2110   Sometimes  Public_Transportation     Obesity_Type_III  \n",
              "\n",
              "[2111 rows x 17 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-73ac5f1e-780d-4916-b690-7506475bb03d\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Height</th>\n",
              "      <th>Weight</th>\n",
              "      <th>family_history_with_overweight</th>\n",
              "      <th>FAVC</th>\n",
              "      <th>FCVC</th>\n",
              "      <th>NCP</th>\n",
              "      <th>CAEC</th>\n",
              "      <th>SMOKE</th>\n",
              "      <th>CH2O</th>\n",
              "      <th>SCC</th>\n",
              "      <th>FAF</th>\n",
              "      <th>TUE</th>\n",
              "      <th>CALC</th>\n",
              "      <th>MTRANS</th>\n",
              "      <th>NObesity</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Female</td>\n",
              "      <td>21.000000</td>\n",
              "      <td>1.620000</td>\n",
              "      <td>64.000000</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>no</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>no</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>no</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>Normal_Weight</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Female</td>\n",
              "      <td>21.000000</td>\n",
              "      <td>1.520000</td>\n",
              "      <td>56.000000</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>yes</td>\n",
              "      <td>3.000000</td>\n",
              "      <td>yes</td>\n",
              "      <td>3.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>Normal_Weight</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Male</td>\n",
              "      <td>23.000000</td>\n",
              "      <td>1.800000</td>\n",
              "      <td>77.000000</td>\n",
              "      <td>yes</td>\n",
              "      <td>no</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>no</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>no</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>Frequently</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>Normal_Weight</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Male</td>\n",
              "      <td>27.000000</td>\n",
              "      <td>1.800000</td>\n",
              "      <td>87.000000</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>no</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>no</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>Frequently</td>\n",
              "      <td>Walking</td>\n",
              "      <td>Overweight_Level_I</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Male</td>\n",
              "      <td>22.000000</td>\n",
              "      <td>1.780000</td>\n",
              "      <td>89.800000</td>\n",
              "      <td>no</td>\n",
              "      <td>no</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>no</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>no</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>Overweight_Level_II</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2106</th>\n",
              "      <td>Female</td>\n",
              "      <td>20.976842</td>\n",
              "      <td>1.710730</td>\n",
              "      <td>131.408528</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>no</td>\n",
              "      <td>1.728139</td>\n",
              "      <td>no</td>\n",
              "      <td>1.676269</td>\n",
              "      <td>0.906247</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>Obesity_Type_III</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2107</th>\n",
              "      <td>Female</td>\n",
              "      <td>21.982942</td>\n",
              "      <td>1.748584</td>\n",
              "      <td>133.742943</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>no</td>\n",
              "      <td>2.005130</td>\n",
              "      <td>no</td>\n",
              "      <td>1.341390</td>\n",
              "      <td>0.599270</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>Obesity_Type_III</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2108</th>\n",
              "      <td>Female</td>\n",
              "      <td>22.524036</td>\n",
              "      <td>1.752206</td>\n",
              "      <td>133.689352</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>no</td>\n",
              "      <td>2.054193</td>\n",
              "      <td>no</td>\n",
              "      <td>1.414209</td>\n",
              "      <td>0.646288</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>Obesity_Type_III</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2109</th>\n",
              "      <td>Female</td>\n",
              "      <td>24.361936</td>\n",
              "      <td>1.739450</td>\n",
              "      <td>133.346641</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>no</td>\n",
              "      <td>2.852339</td>\n",
              "      <td>no</td>\n",
              "      <td>1.139107</td>\n",
              "      <td>0.586035</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>Obesity_Type_III</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2110</th>\n",
              "      <td>Female</td>\n",
              "      <td>23.664709</td>\n",
              "      <td>1.738836</td>\n",
              "      <td>133.472641</td>\n",
              "      <td>yes</td>\n",
              "      <td>yes</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>no</td>\n",
              "      <td>2.863513</td>\n",
              "      <td>no</td>\n",
              "      <td>1.026452</td>\n",
              "      <td>0.714137</td>\n",
              "      <td>Sometimes</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>Obesity_Type_III</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2111 rows × 17 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-73ac5f1e-780d-4916-b690-7506475bb03d')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-73ac5f1e-780d-4916-b690-7506475bb03d button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-73ac5f1e-780d-4916-b690-7506475bb03d');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-11a773c5-8ff4-4733-aa7c-b44ba0dfbbe2\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-11a773c5-8ff4-4733-aa7c-b44ba0dfbbe2')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-11a773c5-8ff4-4733-aa7c-b44ba0dfbbe2 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "dataset",
              "summary": "{\n  \"name\": \"dataset\",\n  \"rows\": 2111,\n  \"fields\": [\n    {\n      \"column\": \"Gender\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"Male\",\n          \"Female\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Age\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 6.345968273732234,\n        \"min\": 14.0,\n        \"max\": 61.0,\n        \"num_unique_values\": 1402,\n        \"samples\": [\n          25.526746,\n          26.740655\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Height\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.09330481986792007,\n        \"min\": 1.45,\n        \"max\": 1.98,\n        \"num_unique_values\": 1574,\n        \"samples\": [\n          1.760175,\n          1.688436\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Weight\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 26.1911717452047,\n        \"min\": 39.0,\n        \"max\": 173.0,\n        \"num_unique_values\": 1525,\n        \"samples\": [\n          120.702935,\n          64.4\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"family_history_with_overweight\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"no\",\n          \"yes\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"FAVC\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"yes\",\n          \"no\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"FCVC\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.5339265785033002,\n        \"min\": 1.0,\n        \"max\": 3.0,\n        \"num_unique_values\": 810,\n        \"samples\": [\n          2.987148,\n          2.939727\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"NCP\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.7780386488418612,\n        \"min\": 1.0,\n        \"max\": 4.0,\n        \"num_unique_values\": 635,\n        \"samples\": [\n          1.468948,\n          2.9948\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"CAEC\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 4,\n        \"samples\": [\n          \"Frequently\",\n          \"no\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"SMOKE\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"yes\",\n          \"no\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"CH2O\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.6129534517968722,\n        \"min\": 1.0,\n        \"max\": 3.0,\n        \"num_unique_values\": 1268,\n        \"samples\": [\n          2.395387,\n          1.983973\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"SCC\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"yes\",\n          \"no\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"FAF\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.850592430836698,\n        \"min\": 0.0,\n        \"max\": 3.0,\n        \"num_unique_values\": 1190,\n        \"samples\": [\n          1.655488,\n          2.433918\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"TUE\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.6089272596763782,\n        \"min\": 0.0,\n        \"max\": 2.0,\n        \"num_unique_values\": 1129,\n        \"samples\": [\n          1.416353,\n          0.878258\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"CALC\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 4,\n        \"samples\": [\n          \"Sometimes\",\n          \"Always\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"MTRANS\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 5,\n        \"samples\": [\n          \"Walking\",\n          \"Bike\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"NObesity\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 7,\n        \"samples\": [\n          \"Normal_Weight\",\n          \"Overweight_Level_I\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 2
        }
      ],
      "source": [
        "dataset"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "n8PpB2-Bxdu9"
      },
      "source": [
        "# Homework specifications\n",
        "\n",
        "\n",
        "Building upon the previous tasks of data preparation, cleaning, and exploratory analysis, you are now required to further your analysis by developing and testing different machine learning algorithms. These algorithms are aimed at predicting specific outcomes based on the dataset. You must implement these algorithms from scratch and may use machine learning frameworks only for verifying the quality of your results.\n",
        "\n",
        "**Algorithms to Implement:**\n",
        "\n",
        "1. **Linear Regression:** To predict the `Weight` of individuals based on the available attributes.\n",
        "2. **Decision Trees for Binary Classification:** To classify individuals into two categories of obesity levels (`NObesity` attribute) based on the available attributes.\n",
        "3. **Logistic Regression:** To classify individuals into two categories of obesity levels (`NObesity` attribute).\n",
        "4. **K-Nearest Neighbors (KNN):** To classify individuals into two categories of obesity levels (`NObesity` attribute).\n",
        "\n",
        "**Data Preparation:**\n",
        "- Categorical attributes defined as strings must be encoded before being used as input to the algorithms. You may choose an appropriate encoding method.\n",
        "- For regression problems, predict the attribute `Weight`.\n",
        "- For classification, predict the attribute `NObesity`. *Binarize the `NObesity` attribute such that categories below and including \"Overweight\" are encoded as `0`, and the rest are encoded as `1`*.\n",
        "\n",
        "**Implementation Guidelines:**\n",
        "- Your implementations must not use any external machine learning frameworks except for verifying the results. For example, you can use scikit-learn to double check your results but not to fit thye model.\n",
        "- External frameworks can be used to do everything except build the model. For instance, you can use Pandas to manage dataframes, or you can use scikit-learn to evaluate the performance of your algorithms.\n",
        "- For each algorithm, you must tune the hyperparameters. Demonstrate where and how you've performed hyperparameter tuning. While exhaustive search is not required, you should clearly indicate the process of hyperparameter tuning and how it can be extended to explore more options.\n",
        "- Ensure that your notebook runs from beginning to end without errors.\n",
        "\n",
        "**Critical Analysis:**\n",
        "- Conclude your notebook with a critical analysis section. This should include observations drawn from your experiments with the implemented algorithms. Discuss the performance, strengths, and weaknesses of each algorithm based on your results.\n",
        "\n",
        "**Hyperparameter Tuning:**\n",
        "- Clearly document the process of tuning hyperparameters for each algorithm. Discuss the impact of different hyperparameters on the algorithm's performance and how further tuning could potentially improve the results.\n",
        "\n",
        "**Submission Requirements:**\n",
        "- Your submission should include the Jupyter Notebook containing all codes for the algorithms, data preparation and cleaning, exploratory data analysis, hyperparameter tuning, and critical analysis.\n",
        "- The notebook must be well-organized, properly commented, and should run from start to finish without any errors.\n",
        "\n",
        "This assignment will challenge your understanding of machine learning concepts, your ability to implement algorithms from scratch, and your analytical skills in interpreting the results. It is an opportunity to demonstrate your proficiency in machine learning and your critical thinking abilities in analyzing and solving problems.\n",
        "\n",
        "**Deadline for Submission:** Before the end of the semester, which is May 31st 2024.\n",
        "\n",
        "Please ensure that your submission is complete and adheres to all the specified requirements. Good luck!\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HlbeKKoQcN2s"
      },
      "source": [
        "##Data Preparation\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "i43A5thadVoy"
      },
      "source": [
        "Funzioni necessarie per trasformare le variabili di tipo stringa in variabili numerali.\n",
        "\n",
        "Per gli attributi MTRANS e NObesity sono presenti due funzioni che eseguono la One Hot Encoding. Per la regressione la One Hot Encoding viene effettiata sia per MTRANS che per NObesity; per la classificazione viene usata solo la One Hot Encoding in riferimento a MTRANS, l'attributo NObesity viene processato diversamente."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "3FJHYIzSoX4r"
      },
      "outputs": [],
      "source": [
        "#Funzioni ausiliarie per trasformare variabili categoriche\n",
        "#in variabili numerali\n",
        "\n",
        "features_to_transform=[\"Gender\", \"family_history_with_overweight\", \"FAVC\", \"SMOKE\", \"SCC\", \"CAEC\", \"CALC\"]\n",
        "\n",
        "\n",
        "def one_hot_transportation(dataset):\n",
        "  dataset[\"mtr\"]=dataset[\"MTRANS\"]\n",
        "  dataset = pd.get_dummies(dataset, columns=[\"MTRANS\"], dtype=int)\n",
        "  dataset = dataset.rename(columns={\"mtr\":\"MTRANS\"})\n",
        "  return dataset\n",
        "\n",
        "def one_hot_obesity(dataset):\n",
        "  dataset[\"ob\"]=dataset[\"NObesity\"]\n",
        "  dataset = pd.get_dummies(dataset, columns=[\"NObesity\"], dtype=int)\n",
        "  dataset = dataset.rename(columns={\"ob\":\"NObesity\"})\n",
        "  return dataset\n",
        "\n",
        "def binary_variable_to_integer(x):\n",
        "  return 1 if (x==\"yes\" or x==\"Male\")  else 0\n",
        "\n",
        "def categorical_value_to_int(x):\n",
        "    if(x==\"Sometimes\"):\n",
        "      return 1\n",
        "    elif(x==\"Frequently\"):\n",
        "      return 2\n",
        "    elif(x==\"Always\"):\n",
        "      return 3\n",
        "    else:\n",
        "      return 0\n",
        "\n",
        "def dataset_to_numerical(dataset, attributes):\n",
        "  ret = dataset.copy()\n",
        "  for i in attributes:\n",
        "    if(i == \"CAEC\" or i==\"CALC\"):\n",
        "      ret[i] = ret[i].apply(categorical_value_to_int)\n",
        "    else:\n",
        "      ret[i] = ret[i].apply(binary_variable_to_integer)\n",
        "  return ret\n",
        "\n",
        "# 0 <= sovrappeso\n",
        "# 1 > sovrappeso\n",
        "\n",
        "def classification_nobesity(value):\n",
        "    if(value== 'Insufficient_Weight' or value=='Normal_Weight' or value==\"Overweight_Level_I\" or value==\"Overweight_Level_II\"):\n",
        "      return 0\n",
        "    else:\n",
        "      return 1\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "RzmvWg546k97"
      },
      "outputs": [],
      "source": [
        "\n",
        "# MinMax normalization\n",
        "# Porta tutte le variabili nel range [0,1]\n",
        "def normalize(dataset, exclude=[]):\n",
        "  normalized = pd.DataFrame()\n",
        "  for col in dataset.columns:\n",
        "    if(col in exclude):\n",
        "      normalized[col] = dataset[col]\n",
        "      continue\n",
        "    max = dataset[col].max()\n",
        "    min = dataset[col].min()\n",
        "    if((max-min)==0):\n",
        "      max = 1\n",
        "    normalized[col] = (dataset[col]-min)/(max-min)\n",
        "\n",
        "  return normalized\n",
        "\n",
        "# Funzione usata per denormalizzare i valori predetti\n",
        "def denormalize(array, min, max):\n",
        "  res = np.array([array]).reshape(-1)\n",
        "  for i in range(len(array)):\n",
        "    res[i]= array[i]*(max-min)+min\n",
        "  return res"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "IJzmzoTC74Rs"
      },
      "outputs": [],
      "source": [
        "# Copiamo il dataset originale, così da poter fare tutte le modifiche necessarie\n",
        "work_dataset = dataset.copy()\n",
        "\n",
        "\n",
        "# One Hot Encoding di MTRANS per l'intero dataset\n",
        "work_dataset = one_hot_transportation(work_dataset)\n",
        "\n",
        "\n",
        "# Trasformazione variabili di tipo stringa -> variabili numerali\n",
        "regression_dataset = dataset_to_numerical(work_dataset.copy(), features_to_transform)\n",
        "\n",
        "\n",
        "# One Hot Encoding NObesity per regressione\n",
        "regression_dataset = one_hot_obesity(regression_dataset)\n",
        "\n",
        "# Trasformazione variabili di tipo stringa -> variabili numerali\n",
        "classification_dataset = dataset_to_numerical(work_dataset.copy(), features_to_transform)\n",
        "\n",
        "# Trasformazione dei valori di NObesity in\n",
        "# 0 <= sovrappeso\n",
        "# 1 > sovrappeso\n",
        "classification_dataset['NObesity'] = classification_dataset['NObesity'].apply(classification_nobesity)\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 444
        },
        "id": "p7xW7dcV-nvR",
        "outputId": "09912322-3a1b-4963-f161-a2e79e176602"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      Gender        Age    Height      Weight  family_history_with_overweight  \\\n",
              "0          0  21.000000  1.620000   64.000000                               1   \n",
              "1          0  21.000000  1.520000   56.000000                               1   \n",
              "2          1  23.000000  1.800000   77.000000                               1   \n",
              "3          1  27.000000  1.800000   87.000000                               0   \n",
              "4          1  22.000000  1.780000   89.800000                               0   \n",
              "...      ...        ...       ...         ...                             ...   \n",
              "2106       0  20.976842  1.710730  131.408528                               1   \n",
              "2107       0  21.982942  1.748584  133.742943                               1   \n",
              "2108       0  22.524036  1.752206  133.689352                               1   \n",
              "2109       0  24.361936  1.739450  133.346641                               1   \n",
              "2110       0  23.664709  1.738836  133.472641                               1   \n",
              "\n",
              "      FAVC  FCVC  NCP  CAEC  SMOKE  ...  MTRANS_Public_Transportation  \\\n",
              "0        0   2.0  3.0     1      0  ...                             1   \n",
              "1        0   3.0  3.0     1      1  ...                             1   \n",
              "2        0   2.0  3.0     1      0  ...                             1   \n",
              "3        0   3.0  3.0     1      0  ...                             0   \n",
              "4        0   2.0  1.0     1      0  ...                             1   \n",
              "...    ...   ...  ...   ...    ...  ...                           ...   \n",
              "2106     1   3.0  3.0     1      0  ...                             1   \n",
              "2107     1   3.0  3.0     1      0  ...                             1   \n",
              "2108     1   3.0  3.0     1      0  ...                             1   \n",
              "2109     1   3.0  3.0     1      0  ...                             1   \n",
              "2110     1   3.0  3.0     1      0  ...                             1   \n",
              "\n",
              "      MTRANS_Walking             NObesity  NObesity_Insufficient_Weight  \\\n",
              "0                  0        Normal_Weight                             0   \n",
              "1                  0        Normal_Weight                             0   \n",
              "2                  0        Normal_Weight                             0   \n",
              "3                  1   Overweight_Level_I                             0   \n",
              "4                  0  Overweight_Level_II                             0   \n",
              "...              ...                  ...                           ...   \n",
              "2106               0     Obesity_Type_III                             0   \n",
              "2107               0     Obesity_Type_III                             0   \n",
              "2108               0     Obesity_Type_III                             0   \n",
              "2109               0     Obesity_Type_III                             0   \n",
              "2110               0     Obesity_Type_III                             0   \n",
              "\n",
              "      NObesity_Normal_Weight NObesity_Obesity_Type_I  \\\n",
              "0                          1                       0   \n",
              "1                          1                       0   \n",
              "2                          1                       0   \n",
              "3                          0                       0   \n",
              "4                          0                       0   \n",
              "...                      ...                     ...   \n",
              "2106                       0                       0   \n",
              "2107                       0                       0   \n",
              "2108                       0                       0   \n",
              "2109                       0                       0   \n",
              "2110                       0                       0   \n",
              "\n",
              "      NObesity_Obesity_Type_II  NObesity_Obesity_Type_III  \\\n",
              "0                            0                          0   \n",
              "1                            0                          0   \n",
              "2                            0                          0   \n",
              "3                            0                          0   \n",
              "4                            0                          0   \n",
              "...                        ...                        ...   \n",
              "2106                         0                          1   \n",
              "2107                         0                          1   \n",
              "2108                         0                          1   \n",
              "2109                         0                          1   \n",
              "2110                         0                          1   \n",
              "\n",
              "      NObesity_Overweight_Level_I  NObesity_Overweight_Level_II  \n",
              "0                               0                             0  \n",
              "1                               0                             0  \n",
              "2                               0                             0  \n",
              "3                               1                             0  \n",
              "4                               0                             1  \n",
              "...                           ...                           ...  \n",
              "2106                            0                             0  \n",
              "2107                            0                             0  \n",
              "2108                            0                             0  \n",
              "2109                            0                             0  \n",
              "2110                            0                             0  \n",
              "\n",
              "[2111 rows x 29 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-a68a9677-799f-4df8-b71f-880ecec6efda\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Height</th>\n",
              "      <th>Weight</th>\n",
              "      <th>family_history_with_overweight</th>\n",
              "      <th>FAVC</th>\n",
              "      <th>FCVC</th>\n",
              "      <th>NCP</th>\n",
              "      <th>CAEC</th>\n",
              "      <th>SMOKE</th>\n",
              "      <th>...</th>\n",
              "      <th>MTRANS_Public_Transportation</th>\n",
              "      <th>MTRANS_Walking</th>\n",
              "      <th>NObesity</th>\n",
              "      <th>NObesity_Insufficient_Weight</th>\n",
              "      <th>NObesity_Normal_Weight</th>\n",
              "      <th>NObesity_Obesity_Type_I</th>\n",
              "      <th>NObesity_Obesity_Type_II</th>\n",
              "      <th>NObesity_Obesity_Type_III</th>\n",
              "      <th>NObesity_Overweight_Level_I</th>\n",
              "      <th>NObesity_Overweight_Level_II</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>21.000000</td>\n",
              "      <td>1.620000</td>\n",
              "      <td>64.000000</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>Normal_Weight</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0</td>\n",
              "      <td>21.000000</td>\n",
              "      <td>1.520000</td>\n",
              "      <td>56.000000</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>Normal_Weight</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>23.000000</td>\n",
              "      <td>1.800000</td>\n",
              "      <td>77.000000</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>Normal_Weight</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1</td>\n",
              "      <td>27.000000</td>\n",
              "      <td>1.800000</td>\n",
              "      <td>87.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>Overweight_Level_I</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1</td>\n",
              "      <td>22.000000</td>\n",
              "      <td>1.780000</td>\n",
              "      <td>89.800000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>Overweight_Level_II</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2106</th>\n",
              "      <td>0</td>\n",
              "      <td>20.976842</td>\n",
              "      <td>1.710730</td>\n",
              "      <td>131.408528</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>Obesity_Type_III</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2107</th>\n",
              "      <td>0</td>\n",
              "      <td>21.982942</td>\n",
              "      <td>1.748584</td>\n",
              "      <td>133.742943</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>Obesity_Type_III</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2108</th>\n",
              "      <td>0</td>\n",
              "      <td>22.524036</td>\n",
              "      <td>1.752206</td>\n",
              "      <td>133.689352</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>Obesity_Type_III</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2109</th>\n",
              "      <td>0</td>\n",
              "      <td>24.361936</td>\n",
              "      <td>1.739450</td>\n",
              "      <td>133.346641</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>Obesity_Type_III</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2110</th>\n",
              "      <td>0</td>\n",
              "      <td>23.664709</td>\n",
              "      <td>1.738836</td>\n",
              "      <td>133.472641</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>Obesity_Type_III</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2111 rows × 29 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a68a9677-799f-4df8-b71f-880ecec6efda')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-a68a9677-799f-4df8-b71f-880ecec6efda button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-a68a9677-799f-4df8-b71f-880ecec6efda');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-96885217-21fe-4d13-86f9-393b722a05e8\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-96885217-21fe-4d13-86f9-393b722a05e8')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-96885217-21fe-4d13-86f9-393b722a05e8 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "regression_dataset"
            }
          },
          "metadata": {},
          "execution_count": 6
        }
      ],
      "source": [
        "regression_dataset"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 444
        },
        "id": "n7304sTH-pcZ",
        "outputId": "bffa40bb-9b41-48d2-bada-f90a1bd3cb40"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      Gender        Age    Height      Weight  family_history_with_overweight  \\\n",
              "0          0  21.000000  1.620000   64.000000                               1   \n",
              "1          0  21.000000  1.520000   56.000000                               1   \n",
              "2          1  23.000000  1.800000   77.000000                               1   \n",
              "3          1  27.000000  1.800000   87.000000                               0   \n",
              "4          1  22.000000  1.780000   89.800000                               0   \n",
              "...      ...        ...       ...         ...                             ...   \n",
              "2106       0  20.976842  1.710730  131.408528                               1   \n",
              "2107       0  21.982942  1.748584  133.742943                               1   \n",
              "2108       0  22.524036  1.752206  133.689352                               1   \n",
              "2109       0  24.361936  1.739450  133.346641                               1   \n",
              "2110       0  23.664709  1.738836  133.472641                               1   \n",
              "\n",
              "      FAVC  FCVC  NCP  CAEC  SMOKE  ...       FAF       TUE  CALC  NObesity  \\\n",
              "0        0   2.0  3.0     1      0  ...  0.000000  1.000000     0         0   \n",
              "1        0   3.0  3.0     1      1  ...  3.000000  0.000000     1         0   \n",
              "2        0   2.0  3.0     1      0  ...  2.000000  1.000000     2         0   \n",
              "3        0   3.0  3.0     1      0  ...  2.000000  0.000000     2         0   \n",
              "4        0   2.0  1.0     1      0  ...  0.000000  0.000000     1         0   \n",
              "...    ...   ...  ...   ...    ...  ...       ...       ...   ...       ...   \n",
              "2106     1   3.0  3.0     1      0  ...  1.676269  0.906247     1         1   \n",
              "2107     1   3.0  3.0     1      0  ...  1.341390  0.599270     1         1   \n",
              "2108     1   3.0  3.0     1      0  ...  1.414209  0.646288     1         1   \n",
              "2109     1   3.0  3.0     1      0  ...  1.139107  0.586035     1         1   \n",
              "2110     1   3.0  3.0     1      0  ...  1.026452  0.714137     1         1   \n",
              "\n",
              "                     MTRANS  MTRANS_Automobile MTRANS_Bike  MTRANS_Motorbike  \\\n",
              "0     Public_Transportation                  0           0                 0   \n",
              "1     Public_Transportation                  0           0                 0   \n",
              "2     Public_Transportation                  0           0                 0   \n",
              "3                   Walking                  0           0                 0   \n",
              "4     Public_Transportation                  0           0                 0   \n",
              "...                     ...                ...         ...               ...   \n",
              "2106  Public_Transportation                  0           0                 0   \n",
              "2107  Public_Transportation                  0           0                 0   \n",
              "2108  Public_Transportation                  0           0                 0   \n",
              "2109  Public_Transportation                  0           0                 0   \n",
              "2110  Public_Transportation                  0           0                 0   \n",
              "\n",
              "      MTRANS_Public_Transportation  MTRANS_Walking  \n",
              "0                                1               0  \n",
              "1                                1               0  \n",
              "2                                1               0  \n",
              "3                                0               1  \n",
              "4                                1               0  \n",
              "...                            ...             ...  \n",
              "2106                             1               0  \n",
              "2107                             1               0  \n",
              "2108                             1               0  \n",
              "2109                             1               0  \n",
              "2110                             1               0  \n",
              "\n",
              "[2111 rows x 22 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-d39564f8-64bf-4e34-8e83-329c0b572e7c\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Height</th>\n",
              "      <th>Weight</th>\n",
              "      <th>family_history_with_overweight</th>\n",
              "      <th>FAVC</th>\n",
              "      <th>FCVC</th>\n",
              "      <th>NCP</th>\n",
              "      <th>CAEC</th>\n",
              "      <th>SMOKE</th>\n",
              "      <th>...</th>\n",
              "      <th>FAF</th>\n",
              "      <th>TUE</th>\n",
              "      <th>CALC</th>\n",
              "      <th>NObesity</th>\n",
              "      <th>MTRANS</th>\n",
              "      <th>MTRANS_Automobile</th>\n",
              "      <th>MTRANS_Bike</th>\n",
              "      <th>MTRANS_Motorbike</th>\n",
              "      <th>MTRANS_Public_Transportation</th>\n",
              "      <th>MTRANS_Walking</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>21.000000</td>\n",
              "      <td>1.620000</td>\n",
              "      <td>64.000000</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0</td>\n",
              "      <td>21.000000</td>\n",
              "      <td>1.520000</td>\n",
              "      <td>56.000000</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>...</td>\n",
              "      <td>3.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>23.000000</td>\n",
              "      <td>1.800000</td>\n",
              "      <td>77.000000</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1</td>\n",
              "      <td>27.000000</td>\n",
              "      <td>1.800000</td>\n",
              "      <td>87.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>Walking</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1</td>\n",
              "      <td>22.000000</td>\n",
              "      <td>1.780000</td>\n",
              "      <td>89.800000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2106</th>\n",
              "      <td>0</td>\n",
              "      <td>20.976842</td>\n",
              "      <td>1.710730</td>\n",
              "      <td>131.408528</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1.676269</td>\n",
              "      <td>0.906247</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2107</th>\n",
              "      <td>0</td>\n",
              "      <td>21.982942</td>\n",
              "      <td>1.748584</td>\n",
              "      <td>133.742943</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1.341390</td>\n",
              "      <td>0.599270</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2108</th>\n",
              "      <td>0</td>\n",
              "      <td>22.524036</td>\n",
              "      <td>1.752206</td>\n",
              "      <td>133.689352</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1.414209</td>\n",
              "      <td>0.646288</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2109</th>\n",
              "      <td>0</td>\n",
              "      <td>24.361936</td>\n",
              "      <td>1.739450</td>\n",
              "      <td>133.346641</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1.139107</td>\n",
              "      <td>0.586035</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2110</th>\n",
              "      <td>0</td>\n",
              "      <td>23.664709</td>\n",
              "      <td>1.738836</td>\n",
              "      <td>133.472641</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1.026452</td>\n",
              "      <td>0.714137</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>Public_Transportation</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2111 rows × 22 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d39564f8-64bf-4e34-8e83-329c0b572e7c')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-d39564f8-64bf-4e34-8e83-329c0b572e7c button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-d39564f8-64bf-4e34-8e83-329c0b572e7c');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-4be09a4b-94ba-4d21-a4ce-5e0355703b8e\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-4be09a4b-94ba-4d21-a4ce-5e0355703b8e')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-4be09a4b-94ba-4d21-a4ce-5e0355703b8e button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "classification_dataset"
            }
          },
          "metadata": {},
          "execution_count": 7
        }
      ],
      "source": [
        "classification_dataset"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vJX_QYLVfEPM"
      },
      "source": [
        "In questa sezione divido i vari dataset in training e test set. Il dataset originale viene inizialmente \"mischiato\" e successivamente si dedica il 90% del dataset alla fase di train e il restante 10% alla la fase di test. <br>\n",
        "Il dataset per la regressione viene normalizzato dopo la sua divisione, questo perché nella fase di test non vi è bisogno di normalizzare le variabili target (Weight), ma nella fase di training è preferibile avere tutte le variabili in uno stesso range, per evitare di dare più peso ad alcune di esse."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "66SSbBVk-vYV"
      },
      "outputs": [],
      "source": [
        "# Con normalizzazione otteniamo tanti valori che poi dovranno\n",
        "# essere discretizzati nel caso dei Decision Tree\n",
        "# quindi copio in anticipo il dataset non normalizzato\n",
        "decision_tree_dataset = classification_dataset.copy()\n",
        "\n",
        "\n",
        "# Normalizzo prima di splittare, non ho bisogno\n",
        "classification_dataset = normalize(classification_dataset, [\"MTRANS\", \"NObesity\"])\n",
        "\n",
        "\n",
        "# DIVIDIAMO IL WORK DATASET IN TRAIN E TEST\n",
        "TRAIN_TEST_SPLIT_PERCENTAGE = 0.9\n",
        "###SHUFFLING\n",
        "regression_dataset = regression_dataset.sample(frac=1, random_state=42).reset_index(drop=True)\n",
        "classification_dataset = classification_dataset.sample(frac=1, random_state=42).reset_index(drop=True)\n",
        "decision_tree_dataset = decision_tree_dataset.sample(frac=1, random_state=42).reset_index(drop=True)\n",
        "\n",
        "# 90% - training set\n",
        "# 10% - test set\n",
        "regression_train = regression_dataset[:int(len(regression_dataset) * TRAIN_TEST_SPLIT_PERCENTAGE)]\n",
        "regression_test = regression_dataset[int(len(regression_dataset) * TRAIN_TEST_SPLIT_PERCENTAGE):]\n",
        "classification_train = classification_dataset[:int(len(classification_dataset) * TRAIN_TEST_SPLIT_PERCENTAGE)]\n",
        "classification_test = classification_dataset[int(len(classification_dataset) * TRAIN_TEST_SPLIT_PERCENTAGE):]\n",
        "\n",
        "\n",
        "# vengono salvati i valori massimi e minimi del target per la regressione\n",
        "# verranno usati poi per denormalizzare\n",
        "r_train_max = regression_train[\"Weight\"].max()\n",
        "r_train_min = regression_train[\"Weight\"].min()\n",
        "\n",
        "regression_test = normalize(regression_test, [\"MTRANS\", \"NObesity\", \"Weight\"])\n",
        "regression_train = normalize(regression_train, [\"MTRANS\", \"NObesity\"])\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4eMVEfwfvnds"
      },
      "outputs": [],
      "source": [
        "\n",
        "# Creazione X train, X test, Y train, Y test per classificazione\n",
        "\n",
        "X_classification_train = classification_train.loc[:, (classification_train.columns != 'NObesity') & (classification_train.columns != 'MTRANS')]\n",
        "Y_classification_train = classification_train[\"NObesity\"]\n",
        "\n",
        "X_classification_test = classification_test.loc[:, (classification_test.columns != 'NObesity') & (classification_test.columns != 'MTRANS')]\n",
        "Y_classification_test = classification_test[\"NObesity\"]\n",
        "\n",
        "# Creazione X train, X test, Y train, Y test per regressione\n",
        "\n",
        "X_regression_train = regression_train.loc[:, (regression_train.columns != 'Weight') & (regression_train.columns != 'MTRANS') & (regression_train.columns != 'NObesity')]\n",
        "Y_regression_train = regression_train[\"Weight\"]\n",
        "\n",
        "X_regression_test = regression_test.loc[:, (regression_test.columns != 'Weight') & (regression_test.columns != 'MTRANS') & (regression_test.columns != 'NObesity')]\n",
        "Y_regression_test = regression_test[\"Weight\"]\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mniLoGEZb8fP"
      },
      "source": [
        "##Grafici Vari"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "PRhINb_JM1I8"
      },
      "outputs": [],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import csv\n",
        "%matplotlib inline"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 565
        },
        "id": "ilT31TNUM1I9",
        "outputId": "43dbd127-7f16-453e-deb4-471492b0b58b"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "plt.figure(figsize=(10, 6))\n",
        "sns.histplot(data=work_dataset, x=\"Age\", kde=True, hue=\"Gender\")\n",
        "plt.title(\"Distribuzione dell\\\"età a seconda del genere\")\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 565
        },
        "id": "rZDM2DRPM1I9",
        "outputId": "bb44e884-69ab-4048-bc6a-79b701366558"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1500x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "plt.figure(figsize=(15, 6))\n",
        "sns.barplot(data=work_dataset, x=\"NObesity\", y=\"Weight\", hue=\"Gender\")\n",
        "plt.title(\"Livello di obesità divisa per peso e genere\")\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 565
        },
        "id": "PeM96ETFM1I-",
        "outputId": "8bf6126d-bc3b-4287-848b-3d5e66a65299"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1500x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "plt.figure(figsize=(15, 6))\n",
        "sns.countplot(data=work_dataset, x=\"MTRANS\", hue=\"NObesity\")\n",
        "plt.title(\"Livello di obesità a seconda del mezzo di trasporto\")\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 564
        },
        "id": "A4P9lKuKM1I-",
        "outputId": "dc9cfe3e-c6b5-4327-b7ee-79185c216960"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1500x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "plt.figure(figsize=(15, 6))\n",
        "sns.countplot(data=work_dataset, x=\"SMOKE\", hue=\"NObesity\")\n",
        "plt.title(\"Livello di obesità a seconda dello status di fumatore\")\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 565
        },
        "id": "SDAOdS7IM1I-",
        "outputId": "b2d59245-2c94-4233-ef49-8128b5ddcf13"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1500x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "plt.figure(figsize=(15, 6))\n",
        "sns.boxplot(data=work_dataset, x=\"NObesity\", y=\"Age\", hue=\"Gender\")\n",
        "plt.title(\"Distribuzione dell'età a seconda del livello di obesità diviso per genere.\")\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "n62_Iy9439Uk"
      },
      "source": [
        "##Linear Regressor"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "SDh5oGaMItfk"
      },
      "outputs": [],
      "source": [
        "from sklearn.metrics import mean_squared_error\n",
        "class LinearRegressor:\n",
        "\n",
        "  def __init__(self, epoch=0, step=0.01, minibatch=0, l2 = 0):\n",
        "    self.epoch=epoch\n",
        "    self.step=step\n",
        "    self.minibatch=minibatch\n",
        "    self.coefficient_value=[]\n",
        "    self.l2=0\n",
        "    self.fitted=False\n",
        "\n",
        "  def _gradientDescent(self, x,y_true,epoch,step, minibatch, l2, printMse):\n",
        "    x_number = x.shape[0]\n",
        "\n",
        "    #inizializzo randomicamente i pesi\n",
        "    weights = np.random.randn(x.shape[1])\n",
        "    i=0\n",
        "\n",
        "    # salvo la trasposta della matrice in input\n",
        "    # così da non doverla caricare ogni volta (solo nel caso in cui non si usi un minibatch)\n",
        "    x_transpose = x.T\n",
        "\n",
        "    while(i<epoch):\n",
        "\n",
        "      if(printMse and i%(epoch/10)==0):\n",
        "        print(\"Epoch: \"+str(i)+\" MSE: \"+str(mean_squared_error( denormalize(y_true, r_train_min, r_train_max), denormalize(np.dot(x, weights), r_train_min, r_train_max))))\n",
        "\n",
        "      # se minibatch!=0 si esegue il gradient descent su un minibatch randomico del dataset\n",
        "      if(minibatch!=0):\n",
        "        permutation = np.random.permutation(x_number)\n",
        "        X_shuffled = x[permutation]\n",
        "        y_shuffled = y_true[permutation]\n",
        "\n",
        "        for j in range(0, x_number, minibatch):\n",
        "          end = j + minibatch\n",
        "\n",
        "          # minibatch di x, y\n",
        "          x_batch = X_shuffled[i:end]\n",
        "          y_batch = y_shuffled[i:end]\n",
        "\n",
        "          # il gradiente è pari a (-2/minibatch_size)*(X.t( Y - pred))\n",
        "          error = y_batch-np.dot(x_batch, weights)\n",
        "          gradient = (-2/minibatch)*np.dot(x_batch.T, error)\n",
        "\n",
        "          #formula di aggiornamento pesi\n",
        "          weights = weights - step*(gradient + l2*weights)\n",
        "\n",
        "      else:\n",
        "        # gradient descent su intero dataset\n",
        "        error = y_true-np.dot(x, weights)\n",
        "        gradient = (-2/x_number)*np.dot(x_transpose, error)\n",
        "        weights = weights - step*( gradient + l2*weights)\n",
        "      i+=1\n",
        "    return weights\n",
        "\n",
        "\n",
        "\n",
        "  def fit(self, x, y_true, printMse=True):\n",
        "\n",
        "    # si inserisce una colonna dummy fatta di soli 1\n",
        "    # che rappresenta il bias\n",
        "    if 'Bias' not in x.columns:\n",
        "      x.insert(0, column=\"Bias\", value=1)\n",
        "    x_t = np.array(x)\n",
        "    y = np.array(y_true)\n",
        "    # Se non sono specificate le epoche, si usa la Normal Equation\n",
        "    if(self.epoch == 0):\n",
        "      self.coefficient_value = np.dot(np.linalg.pinv(x_t), y);\n",
        "    else:\n",
        "      self.coefficient_value =  self._gradientDescent(x_t,y, self.epoch, self.step, self.minibatch, self.l2, printMse)\n",
        "    self.fitted=True\n",
        "\n",
        "\n",
        "  def predict(self, x):\n",
        "    if not self.fitted:\n",
        "      raise Exception(\"Fit the model first\")\n",
        "    if 'Bias' not in x.columns:\n",
        "      x.insert(0, column=\"Bias\", value=1)\n",
        "\n",
        "\n",
        "    # Pred = X*W\n",
        "    res = np.dot(x, self.coefficient_value)\n",
        "    return denormalize(res, r_train_min, r_train_max)\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ElTPbEZ7HP-r"
      },
      "source": [
        "##Logistic Classifier\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "SXuVinbr_C4y"
      },
      "outputs": [],
      "source": [
        "from sklearn.metrics import accuracy_score\n",
        "\n",
        "\n",
        "class LogisticClassifier:\n",
        "\n",
        "  def __init__(self, epoch=10000, step=0.01, l2=0, threshold=0.5):\n",
        "    self.epoch = epoch\n",
        "    self.step = step\n",
        "    self.threshold = threshold\n",
        "    self.coefficient_value = []\n",
        "    self.l2 = 0\n",
        "    self.fitted = False\n",
        "\n",
        "  def _sigmoid(self, x):\n",
        "    return 1 / (1 + np.exp(-x))\n",
        "\n",
        "\n",
        "  def fit(self, x, y_true, printAccuracy=True):\n",
        "\n",
        "    # Colonna dummy di soli 1 che rappresenta il bias\n",
        "    if 'Bias' not in x.columns:\n",
        "      x.insert(0, column=\"Bias\", value=1)\n",
        "    x_transpose = x.T\n",
        "    weights = np.random.randn(x.shape[1])\n",
        "\n",
        "    for i in range(self.epoch):\n",
        "      if(printAccuracy and i%(self.epoch/10)==0):\n",
        "\n",
        "        # predizione lineare\n",
        "        y_hat = np.dot(x, weights)\n",
        "\n",
        "        # viene passata nella sigmoide\n",
        "        y_predicted = self._sigmoid(y_hat)\n",
        "\n",
        "        #se y_predicted è maggiore della threshold allora è 1, altrimenti 0\n",
        "        result = [1 if i > self.threshold else 0 for i in y_predicted]\n",
        "\n",
        "        print(\"Epoch: \"+str(i)+\" Accuracy: \"+str(accuracy_score(result, y_true)))\n",
        "\n",
        "      # predizione lineare = X*W\n",
        "      linear_prediction = np.dot(x, weights)\n",
        "\n",
        "      # si passa la predizione lineare nella sigmoide\n",
        "      prediction = self._sigmoid(linear_prediction)\n",
        "\n",
        "      error = y_true - prediction\n",
        "\n",
        "      # gradiente = -2/dataset_size*(X.t*(Y-y_pred))\n",
        "      gradient = (-2 / x.shape[0]) * np.dot(x_transpose, error)\n",
        "\n",
        "      #aggiornamento pesi\n",
        "      weights -= self.step *(gradient) + self.step*(self.l2*weights)\n",
        "\n",
        "    self.fitted = True\n",
        "    self.coefficient_value=weights\n",
        "\n",
        "  def predict(self, x):\n",
        "    if(not self.fitted):\n",
        "      raise Exception(\"Fit the model\")\n",
        "    if 'Bias' not in x.columns:\n",
        "      x.insert(0, column=\"Bias\", value=1)\n",
        "    y_hat = np.dot(x, self.coefficient_value)\n",
        "    y_predicted = self._sigmoid(y_hat)\n",
        "\n",
        "    # 1 se la predizione supera la threshold\n",
        "    # 0 altrimenti\n",
        "    result = [1 if i > self.threshold else 0 for i in y_predicted]\n",
        "\n",
        "    return np.array(result)\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "upPOeNcZ3wG_"
      },
      "source": [
        "##K-Nearest-Neighbour Classifier"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4rnTaTCbqPKo"
      },
      "outputs": [],
      "source": [
        "from collections import Counter\n",
        "class KNNClassifier:\n",
        "\n",
        "  def __init__(self, k=3, distance_type=2):\n",
        "    self.k = k\n",
        "    self.distance_type = distance_type\n",
        "    self.fitted = False\n",
        "\n",
        "\n",
        "  # calcola distanza l2\n",
        "  def _euclidean_distance(self,x1,x2):\n",
        "    return np.linalg.norm(x1-x2)\n",
        "\n",
        "  #calcola distanza l1\n",
        "  def _manhattan_distance(self, x1,x2):\n",
        "    return np.linalg.norm(x1-x2, 1)\n",
        "\n",
        "\n",
        "  # prende in input una riga di un dataset\n",
        "  # e ne calcola la distanza da tutte le righe\n",
        "  # del dataset dataset di train\n",
        "  def _compute_nearest(self, x):\n",
        "    distance = []\n",
        "    for x_train in self.X_train:\n",
        "      if(self.distance_type == 2):\n",
        "        distance.append(self._euclidean_distance(x, x_train))\n",
        "      else:\n",
        "        distance.append(self._manhattan_distance(x, x_train))\n",
        "\n",
        "    # Ordina le distanze, da minore a maggiore e salva i corrispettivi indici\n",
        "    # tra questi vado a prendere i primi k\n",
        "    k_nearest_indices = np.argsort(distance)[:self.k]\n",
        "    k_nearest_elements = self.Y_train[k_nearest_indices]\n",
        "\n",
        "\n",
        "    # Prendo il valore con più occorrenze tra i k e lo fornisco in output\n",
        "    most_common = Counter(k_nearest_elements).most_common(1)[0][0]\n",
        "\n",
        "    return most_common\n",
        "\n",
        "\n",
        "  def fit(self, x, y_true):\n",
        "    if(\"Bias\" in x.columns):\n",
        "      self.X_train = x.drop(columns=[\"Bias\"]).to_numpy()\n",
        "    else:\n",
        "      self.X_train = x.to_numpy()\n",
        "\n",
        "    self.Y_train = y_true.to_numpy()\n",
        "\n",
        "    if((self.k)%2==0):\n",
        "      raise Exception(\"Insert an odd number for k\")\n",
        "    if(self.distance_type != 1 and self.distance_type!=2):\n",
        "      raise Exception(\"Insert a valid distance metric (1 - 2)\")\n",
        "    self.fitted = True\n",
        "\n",
        "  def predict(self, X):\n",
        "    if(\"Bias\" in X.columns):\n",
        "      X_t = X.drop(columns=[\"Bias\"]).to_numpy()\n",
        "    else:\n",
        "      X_t = X.to_numpy()\n",
        "    if(not self.fitted):\n",
        "      raise Exception(\"Fit the model first\")\n",
        "\n",
        "    # passo di volta in volta le righe dell'input X\n",
        "    # alla funzione compute_nearest\n",
        "\n",
        "    predict = [self._compute_nearest(x) for x in X_t]\n",
        "    return np.array(predict)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TqyRcPM9VX_n"
      },
      "source": [
        "Confronto con KNN Scikit learn"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "U2v8akrT0-Qj"
      },
      "source": [
        "Prima di creare il modello di Decision Tree devo attuare delle modifiche sul dataset. Le colonne \"Weight\", \"Height\" e \"Age\" sono continue, abbiamo bisogno di discretizzarle per evitare che il Decision Tree crei troppi rami in base a questi valori.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iJu4_qmH_KRQ"
      },
      "source": [
        "##Decision Tree"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EEu62UcomfpJ"
      },
      "source": [
        "Per i Decision Tree vi è un ulteriore passo nel Data Preparation. Tutte le variabili continue (es. Weight) vengono arrotondate in modo da ridurre il numero di possibili split scelti dall'albero.\n",
        "<br>\n",
        "<br>\n",
        "Un'altra tecnica che si sarebbe potuta utilizzare è quella del \"Binning\" in cui le variabili continue verrebbero raggruppate in sottocategorie, i \"bins\"."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 444
        },
        "id": "xWQy2BGVo6l6",
        "outputId": "489feabe-42de-4797-8003-96f19803aa19"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      Gender   Age  Height  Weight  family_history_with_overweight  FAVC  \\\n",
              "0          0  20.4    1.76      53                               1     1   \n",
              "1          0  26.0    1.62     111                               1     1   \n",
              "2          1  18.0    1.85      60                               1     1   \n",
              "3          0  21.0    1.52      42                               0     1   \n",
              "4          1  22.0    1.75      74                               1     0   \n",
              "...      ...   ...     ...     ...                             ...   ...   \n",
              "2106       1  32.3    1.75     120                               1     1   \n",
              "2107       1  23.0    1.72      81                               1     1   \n",
              "2108       0  23.0    1.65      80                               1     1   \n",
              "2109       0  23.0    1.63      84                               1     1   \n",
              "2110       1  20.0    1.82      85                               1     1   \n",
              "\n",
              "      FCVC  NCP  CAEC  SMOKE  ...  SCC  FAF  TUE  CALC  NObesity  \\\n",
              "0      2.0  3.9     2      0  ...    0  2.9  2.0     0         0   \n",
              "1      3.0  3.0     1      0  ...    0  0.0  0.3     1         1   \n",
              "2      3.0  4.0     1      0  ...    1  2.0  0.0     1         0   \n",
              "3      3.0  1.0     2      0  ...    0  0.0  0.0     1         0   \n",
              "4      2.0  3.0     1      0  ...    0  1.0  2.0     1         0   \n",
              "...    ...  ...   ...    ...  ...  ...  ...  ...   ...       ...   \n",
              "2106   3.0  3.0     1      0  ...    0  1.0  1.3     1         1   \n",
              "2107   2.0  1.7     1      0  ...    0  0.9  1.0     1         0   \n",
              "2108   2.0  3.0     1      0  ...    0  0.1  2.0     0         0   \n",
              "2109   2.1  3.0     1      0  ...    0  0.9  0.6     0         1   \n",
              "2110   3.0  3.0     1      0  ...    0  2.4  0.6     1         0   \n",
              "\n",
              "      MTRANS_Automobile  MTRANS_Bike  MTRANS_Motorbike  \\\n",
              "0                     0            0                 0   \n",
              "1                     0            0                 0   \n",
              "2                     1            0                 0   \n",
              "3                     0            0                 0   \n",
              "4                     0            1                 0   \n",
              "...                 ...          ...               ...   \n",
              "2106                  1            0                 0   \n",
              "2107                  0            0                 0   \n",
              "2108                  0            0                 0   \n",
              "2109                  0            0                 0   \n",
              "2110                  0            0                 0   \n",
              "\n",
              "      MTRANS_Public_Transportation  MTRANS_Walking  \n",
              "0                                1               0  \n",
              "1                                1               0  \n",
              "2                                0               0  \n",
              "3                                1               0  \n",
              "4                                0               0  \n",
              "...                            ...             ...  \n",
              "2106                             0               0  \n",
              "2107                             1               0  \n",
              "2108                             1               0  \n",
              "2109                             1               0  \n",
              "2110                             1               0  \n",
              "\n",
              "[2111 rows x 21 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-150156f3-8cba-4e25-a553-7cb0fed8e109\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Height</th>\n",
              "      <th>Weight</th>\n",
              "      <th>family_history_with_overweight</th>\n",
              "      <th>FAVC</th>\n",
              "      <th>FCVC</th>\n",
              "      <th>NCP</th>\n",
              "      <th>CAEC</th>\n",
              "      <th>SMOKE</th>\n",
              "      <th>...</th>\n",
              "      <th>SCC</th>\n",
              "      <th>FAF</th>\n",
              "      <th>TUE</th>\n",
              "      <th>CALC</th>\n",
              "      <th>NObesity</th>\n",
              "      <th>MTRANS_Automobile</th>\n",
              "      <th>MTRANS_Bike</th>\n",
              "      <th>MTRANS_Motorbike</th>\n",
              "      <th>MTRANS_Public_Transportation</th>\n",
              "      <th>MTRANS_Walking</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>20.4</td>\n",
              "      <td>1.76</td>\n",
              "      <td>53</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.9</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>2.9</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0</td>\n",
              "      <td>26.0</td>\n",
              "      <td>1.62</td>\n",
              "      <td>111</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>18.0</td>\n",
              "      <td>1.85</td>\n",
              "      <td>60</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0</td>\n",
              "      <td>21.0</td>\n",
              "      <td>1.52</td>\n",
              "      <td>42</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1</td>\n",
              "      <td>22.0</td>\n",
              "      <td>1.75</td>\n",
              "      <td>74</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2106</th>\n",
              "      <td>1</td>\n",
              "      <td>32.3</td>\n",
              "      <td>1.75</td>\n",
              "      <td>120</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2107</th>\n",
              "      <td>1</td>\n",
              "      <td>23.0</td>\n",
              "      <td>1.72</td>\n",
              "      <td>81</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.7</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0.9</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2108</th>\n",
              "      <td>0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>1.65</td>\n",
              "      <td>80</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0.1</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2109</th>\n",
              "      <td>0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>1.63</td>\n",
              "      <td>84</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2.1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0.9</td>\n",
              "      <td>0.6</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2110</th>\n",
              "      <td>1</td>\n",
              "      <td>20.0</td>\n",
              "      <td>1.82</td>\n",
              "      <td>85</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>2.4</td>\n",
              "      <td>0.6</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2111 rows × 21 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-150156f3-8cba-4e25-a553-7cb0fed8e109')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-150156f3-8cba-4e25-a553-7cb0fed8e109 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-150156f3-8cba-4e25-a553-7cb0fed8e109');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-a549bc04-636c-4eba-a6b1-489bbadd8a78\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-a549bc04-636c-4eba-a6b1-489bbadd8a78')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-a549bc04-636c-4eba-a6b1-489bbadd8a78 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "decision_tree_dataset"
            }
          },
          "metadata": {},
          "execution_count": 19
        }
      ],
      "source": [
        "decision_tree_dataset = decision_tree_dataset.drop(columns=[\"MTRANS\"])\n",
        "for i in decision_tree_dataset.columns:\n",
        "  if(i == \"Weight\"):\n",
        "    decision_tree_dataset[i] = decision_tree_dataset[i].apply(lambda x: int(x))\n",
        "  elif(i == \"Height\"):\n",
        "    decision_tree_dataset[i] = decision_tree_dataset[i].apply(lambda x: round(x, 2))\n",
        "  else:\n",
        "    decision_tree_dataset[i] = decision_tree_dataset[i].apply(lambda x: round(x, 1))\n",
        "\n",
        "decision_tree_dataset"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rIcnxPMA4clK"
      },
      "source": [
        "Esempio binning"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 444
        },
        "id": "f3qKt75a75X0",
        "outputId": "590f0f37-876e-4d2f-a2db-66938087ab09"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      Gender    Age        Height        Weight  \\\n",
              "0          0  Young  MediumHeight     LowWeight   \n",
              "1          0  Young  MediumHeight    HighWeight   \n",
              "2          1  Young    HighHeight     LowWeight   \n",
              "3          0  Young     LowHeight     LowWeight   \n",
              "4          1  Young  MediumHeight  MediumWeight   \n",
              "...      ...    ...           ...           ...   \n",
              "2106       1  Adult  MediumHeight    HighWeight   \n",
              "2107       1  Young  MediumHeight  MediumWeight   \n",
              "2108       0  Young  MediumHeight  MediumWeight   \n",
              "2109       0  Young  MediumHeight  MediumWeight   \n",
              "2110       1  Young    HighHeight  MediumWeight   \n",
              "\n",
              "      family_history_with_overweight  FAVC             FCVC        NCP  CAEC  \\\n",
              "0                                  1     1  MediumFrequency  ManyMeals     2   \n",
              "1                                  1     1    HighFrequency  ManyMeals     1   \n",
              "2                                  1     1    HighFrequency  ManyMeals     1   \n",
              "3                                  0     1    HighFrequency   FewMeals     2   \n",
              "4                                  1     0  MediumFrequency  ManyMeals     1   \n",
              "...                              ...   ...              ...        ...   ...   \n",
              "2106                               1     1    HighFrequency  ManyMeals     1   \n",
              "2107                               1     1  MediumFrequency   FewMeals     1   \n",
              "2108                               1     1  MediumFrequency  ManyMeals     1   \n",
              "2109                               1     1  MediumFrequency  ManyMeals     1   \n",
              "2110                               1     1    HighFrequency  ManyMeals     1   \n",
              "\n",
              "      SMOKE                  CH2O  SCC              FAF  \\\n",
              "0         0  MediumH2OConsumption    0    HighFrequency   \n",
              "1         0    HighH2OConsumption    0     LowFrequency   \n",
              "2         0  MediumH2OConsumption    1    HighFrequency   \n",
              "3         0     LowH2OConsumption    0     LowFrequency   \n",
              "4         0  MediumH2OConsumption    0  MediumFrequency   \n",
              "...     ...                   ...  ...              ...   \n",
              "2106      0    HighH2OConsumption    0  MediumFrequency   \n",
              "2107      0     LowH2OConsumption    0  MediumFrequency   \n",
              "2108      0  MediumH2OConsumption    0     LowFrequency   \n",
              "2109      0  MediumH2OConsumption    0  MediumFrequency   \n",
              "2110      0    HighH2OConsumption    0    HighFrequency   \n",
              "\n",
              "                        TUE  CALC  NObesity  \n",
              "0       HighTechnologyUsage     0         0  \n",
              "1        LowTechnologyUsage     1         1  \n",
              "2        LowTechnologyUsage     1         0  \n",
              "3        LowTechnologyUsage     1         0  \n",
              "4       HighTechnologyUsage     1         0  \n",
              "...                     ...   ...       ...  \n",
              "2106  MediumTechnologyUsage     1         1  \n",
              "2107  MediumTechnologyUsage     1         0  \n",
              "2108    HighTechnologyUsage     0         0  \n",
              "2109     LowTechnologyUsage     0         1  \n",
              "2110     LowTechnologyUsage     1         0  \n",
              "\n",
              "[2111 rows x 16 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1b9246f2-9a49-43ab-ae10-3580f47704d4\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Height</th>\n",
              "      <th>Weight</th>\n",
              "      <th>family_history_with_overweight</th>\n",
              "      <th>FAVC</th>\n",
              "      <th>FCVC</th>\n",
              "      <th>NCP</th>\n",
              "      <th>CAEC</th>\n",
              "      <th>SMOKE</th>\n",
              "      <th>CH2O</th>\n",
              "      <th>SCC</th>\n",
              "      <th>FAF</th>\n",
              "      <th>TUE</th>\n",
              "      <th>CALC</th>\n",
              "      <th>NObesity</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>Young</td>\n",
              "      <td>MediumHeight</td>\n",
              "      <td>LowWeight</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>MediumFrequency</td>\n",
              "      <td>ManyMeals</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>MediumH2OConsumption</td>\n",
              "      <td>0</td>\n",
              "      <td>HighFrequency</td>\n",
              "      <td>HighTechnologyUsage</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0</td>\n",
              "      <td>Young</td>\n",
              "      <td>MediumHeight</td>\n",
              "      <td>HighWeight</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>HighFrequency</td>\n",
              "      <td>ManyMeals</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>HighH2OConsumption</td>\n",
              "      <td>0</td>\n",
              "      <td>LowFrequency</td>\n",
              "      <td>LowTechnologyUsage</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>Young</td>\n",
              "      <td>HighHeight</td>\n",
              "      <td>LowWeight</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>HighFrequency</td>\n",
              "      <td>ManyMeals</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>MediumH2OConsumption</td>\n",
              "      <td>1</td>\n",
              "      <td>HighFrequency</td>\n",
              "      <td>LowTechnologyUsage</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0</td>\n",
              "      <td>Young</td>\n",
              "      <td>LowHeight</td>\n",
              "      <td>LowWeight</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>HighFrequency</td>\n",
              "      <td>FewMeals</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>LowH2OConsumption</td>\n",
              "      <td>0</td>\n",
              "      <td>LowFrequency</td>\n",
              "      <td>LowTechnologyUsage</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1</td>\n",
              "      <td>Young</td>\n",
              "      <td>MediumHeight</td>\n",
              "      <td>MediumWeight</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>MediumFrequency</td>\n",
              "      <td>ManyMeals</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>MediumH2OConsumption</td>\n",
              "      <td>0</td>\n",
              "      <td>MediumFrequency</td>\n",
              "      <td>HighTechnologyUsage</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2106</th>\n",
              "      <td>1</td>\n",
              "      <td>Adult</td>\n",
              "      <td>MediumHeight</td>\n",
              "      <td>HighWeight</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>HighFrequency</td>\n",
              "      <td>ManyMeals</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>HighH2OConsumption</td>\n",
              "      <td>0</td>\n",
              "      <td>MediumFrequency</td>\n",
              "      <td>MediumTechnologyUsage</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2107</th>\n",
              "      <td>1</td>\n",
              "      <td>Young</td>\n",
              "      <td>MediumHeight</td>\n",
              "      <td>MediumWeight</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>MediumFrequency</td>\n",
              "      <td>FewMeals</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>LowH2OConsumption</td>\n",
              "      <td>0</td>\n",
              "      <td>MediumFrequency</td>\n",
              "      <td>MediumTechnologyUsage</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2108</th>\n",
              "      <td>0</td>\n",
              "      <td>Young</td>\n",
              "      <td>MediumHeight</td>\n",
              "      <td>MediumWeight</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>MediumFrequency</td>\n",
              "      <td>ManyMeals</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>MediumH2OConsumption</td>\n",
              "      <td>0</td>\n",
              "      <td>LowFrequency</td>\n",
              "      <td>HighTechnologyUsage</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2109</th>\n",
              "      <td>0</td>\n",
              "      <td>Young</td>\n",
              "      <td>MediumHeight</td>\n",
              "      <td>MediumWeight</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>MediumFrequency</td>\n",
              "      <td>ManyMeals</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>MediumH2OConsumption</td>\n",
              "      <td>0</td>\n",
              "      <td>MediumFrequency</td>\n",
              "      <td>LowTechnologyUsage</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2110</th>\n",
              "      <td>1</td>\n",
              "      <td>Young</td>\n",
              "      <td>HighHeight</td>\n",
              "      <td>MediumWeight</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>HighFrequency</td>\n",
              "      <td>ManyMeals</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>HighH2OConsumption</td>\n",
              "      <td>0</td>\n",
              "      <td>HighFrequency</td>\n",
              "      <td>LowTechnologyUsage</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2111 rows × 16 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1b9246f2-9a49-43ab-ae10-3580f47704d4')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-1b9246f2-9a49-43ab-ae10-3580f47704d4 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-1b9246f2-9a49-43ab-ae10-3580f47704d4');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-8e2fe88e-f070-4c56-b107-32f1cae98505\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-8e2fe88e-f070-4c56-b107-32f1cae98505')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-8e2fe88e-f070-4c56-b107-32f1cae98505 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "decision_tree_dataset_binning",
              "summary": "{\n  \"name\": \"decision_tree_dataset_binning\",\n  \"rows\": 2111,\n  \"fields\": [\n    {\n      \"column\": \"Gender\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 1,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Age\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          \"Young\",\n          \"Adult\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Height\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          \"MediumHeight\",\n          \"HighHeight\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Weight\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          \"LowWeight\",\n          \"HighWeight\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"family_history_with_overweight\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 1,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          0,\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"FAVC\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 1,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          0,\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"FCVC\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          \"MediumFrequency\",\n          \"HighFrequency\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"NCP\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          \"ManyMeals\",\n          \"FewMeals\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"CAEC\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 3,\n        \"num_unique_values\": 4,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"SMOKE\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 1,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"CH2O\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          \"MediumH2OConsumption\",\n          \"HighH2OConsumption\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"SCC\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 1,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"FAF\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          \"HighFrequency\",\n          \"LowFrequency\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"TUE\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          \"HighTechnologyUsage\",\n          \"LowTechnologyUsage\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"CALC\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 3,\n        \"num_unique_values\": 4,\n        \"samples\": [\n          1,\n          3\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"NObesity\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 1,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 20
        }
      ],
      "source": [
        "decision_tree_dataset_binning = decision_tree_dataset.drop(columns=['MTRANS_Automobile', 'MTRANS_Bike', 'MTRANS_Motorbike', 'MTRANS_Public_Transportation', 'MTRANS_Walking'])\n",
        "decision_tree_dataset_binning[\"Age\"] = pd.cut(decision_tree_dataset_binning[\"Age\"], bins=[12,30,60,100], labels=[\"Young\", \"Adult\", \"Old\"],right=True, include_lowest=True)\n",
        "decision_tree_dataset_binning[\"Weight\"] = pd.cut(decision_tree_dataset_binning[\"Weight\"], bins=[29,60,90,200], labels=[\"LowWeight\", \"MediumWeight\", \"HighWeight\"],right=True, include_lowest=True)\n",
        "decision_tree_dataset_binning[\"Height\"] = pd.cut(decision_tree_dataset_binning[\"Height\"], bins=[1.30,1.60, 1.80, 2.5], labels=[\"LowHeight\", \"MediumHeight\", \"HighHeight\"],right=True, include_lowest=True)\n",
        "decision_tree_dataset_binning[\"FAF\"] = pd.cut(decision_tree_dataset_binning[\"FAF\"], bins=[0,0.75, 1.5, 3.5], labels=[\"LowFrequency\", \"MediumFrequency\", \"HighFrequency\"],right=True, include_lowest=True)\n",
        "decision_tree_dataset_binning[\"NCP\"] = pd.cut(decision_tree_dataset_binning[\"NCP\"], bins=[0,1.75, 2.5, 4.5], labels=[\"FewMeals\", \"NormalMeals\", \"ManyMeals\"],right=True, include_lowest=True)\n",
        "decision_tree_dataset_binning[\"FCVC\"] = pd.cut(decision_tree_dataset_binning[\"FCVC\"], bins=[0,1.75, 2.25, 3.5], labels=[\"LowFrequency\", \"MediumFrequency\", \"HighFrequency\"],right=True, include_lowest=True)\n",
        "decision_tree_dataset_binning[\"TUE\"] = pd.cut(decision_tree_dataset_binning[\"TUE\"], bins=[0,0.75, 1.50, 2.5], labels=[\"LowTechnologyUsage\", \"MediumTechnologyUsage\", \"HighTechnologyUsage\"], right=True, include_lowest=True)\n",
        "decision_tree_dataset_binning[\"CH2O\"] = pd.cut(decision_tree_dataset_binning[\"CH2O\"], bins=[0.75, 1.50,2.25, 3], labels=[\"LowH2OConsumption\", \"MediumH2OConsumption\", \"HighH2OConsumption\"], right=True, include_lowest=True)\n",
        "decision_tree_dataset_binning"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 444
        },
        "id": "z5whhaKQpxxp",
        "outputId": "f3575251-5d6d-48b0-c070-62454271c935"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      Gender   Age  Height  Weight  family_history_with_overweight  FAVC  \\\n",
              "0          0  20.4    1.76      53                               1     1   \n",
              "1          0  26.0    1.62     111                               1     1   \n",
              "2          1  18.0    1.85      60                               1     1   \n",
              "3          0  21.0    1.52      42                               0     1   \n",
              "4          1  22.0    1.75      74                               1     0   \n",
              "...      ...   ...     ...     ...                             ...   ...   \n",
              "2106       1  32.3    1.75     120                               1     1   \n",
              "2107       1  23.0    1.72      81                               1     1   \n",
              "2108       0  23.0    1.65      80                               1     1   \n",
              "2109       0  23.0    1.63      84                               1     1   \n",
              "2110       1  20.0    1.82      85                               1     1   \n",
              "\n",
              "      FCVC  NCP  CAEC  SMOKE  ...  SCC  FAF  TUE  CALC  NObesity  \\\n",
              "0      2.0  3.9     2      0  ...    0  2.9  2.0     0         0   \n",
              "1      3.0  3.0     1      0  ...    0  0.0  0.3     1         1   \n",
              "2      3.0  4.0     1      0  ...    1  2.0  0.0     1         0   \n",
              "3      3.0  1.0     2      0  ...    0  0.0  0.0     1         0   \n",
              "4      2.0  3.0     1      0  ...    0  1.0  2.0     1         0   \n",
              "...    ...  ...   ...    ...  ...  ...  ...  ...   ...       ...   \n",
              "2106   3.0  3.0     1      0  ...    0  1.0  1.3     1         1   \n",
              "2107   2.0  1.7     1      0  ...    0  0.9  1.0     1         0   \n",
              "2108   2.0  3.0     1      0  ...    0  0.1  2.0     0         0   \n",
              "2109   2.1  3.0     1      0  ...    0  0.9  0.6     0         1   \n",
              "2110   3.0  3.0     1      0  ...    0  2.4  0.6     1         0   \n",
              "\n",
              "      MTRANS_Automobile  MTRANS_Bike  MTRANS_Motorbike  \\\n",
              "0                     0            0                 0   \n",
              "1                     0            0                 0   \n",
              "2                     1            0                 0   \n",
              "3                     0            0                 0   \n",
              "4                     0            1                 0   \n",
              "...                 ...          ...               ...   \n",
              "2106                  1            0                 0   \n",
              "2107                  0            0                 0   \n",
              "2108                  0            0                 0   \n",
              "2109                  0            0                 0   \n",
              "2110                  0            0                 0   \n",
              "\n",
              "      MTRANS_Public_Transportation  MTRANS_Walking  \n",
              "0                                1               0  \n",
              "1                                1               0  \n",
              "2                                0               0  \n",
              "3                                1               0  \n",
              "4                                0               0  \n",
              "...                            ...             ...  \n",
              "2106                             0               0  \n",
              "2107                             1               0  \n",
              "2108                             1               0  \n",
              "2109                             1               0  \n",
              "2110                             1               0  \n",
              "\n",
              "[2111 rows x 21 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-0d2fa3c5-3e82-4b5c-95bf-d43b90b1127e\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Height</th>\n",
              "      <th>Weight</th>\n",
              "      <th>family_history_with_overweight</th>\n",
              "      <th>FAVC</th>\n",
              "      <th>FCVC</th>\n",
              "      <th>NCP</th>\n",
              "      <th>CAEC</th>\n",
              "      <th>SMOKE</th>\n",
              "      <th>...</th>\n",
              "      <th>SCC</th>\n",
              "      <th>FAF</th>\n",
              "      <th>TUE</th>\n",
              "      <th>CALC</th>\n",
              "      <th>NObesity</th>\n",
              "      <th>MTRANS_Automobile</th>\n",
              "      <th>MTRANS_Bike</th>\n",
              "      <th>MTRANS_Motorbike</th>\n",
              "      <th>MTRANS_Public_Transportation</th>\n",
              "      <th>MTRANS_Walking</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>20.4</td>\n",
              "      <td>1.76</td>\n",
              "      <td>53</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.9</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>2.9</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0</td>\n",
              "      <td>26.0</td>\n",
              "      <td>1.62</td>\n",
              "      <td>111</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>18.0</td>\n",
              "      <td>1.85</td>\n",
              "      <td>60</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>1</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0</td>\n",
              "      <td>21.0</td>\n",
              "      <td>1.52</td>\n",
              "      <td>42</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1</td>\n",
              "      <td>22.0</td>\n",
              "      <td>1.75</td>\n",
              "      <td>74</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2106</th>\n",
              "      <td>1</td>\n",
              "      <td>32.3</td>\n",
              "      <td>1.75</td>\n",
              "      <td>120</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2107</th>\n",
              "      <td>1</td>\n",
              "      <td>23.0</td>\n",
              "      <td>1.72</td>\n",
              "      <td>81</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.7</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0.9</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2108</th>\n",
              "      <td>0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>1.65</td>\n",
              "      <td>80</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0.1</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2109</th>\n",
              "      <td>0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>1.63</td>\n",
              "      <td>84</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2.1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0.9</td>\n",
              "      <td>0.6</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2110</th>\n",
              "      <td>1</td>\n",
              "      <td>20.0</td>\n",
              "      <td>1.82</td>\n",
              "      <td>85</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>2.4</td>\n",
              "      <td>0.6</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2111 rows × 21 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0d2fa3c5-3e82-4b5c-95bf-d43b90b1127e')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-0d2fa3c5-3e82-4b5c-95bf-d43b90b1127e button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-0d2fa3c5-3e82-4b5c-95bf-d43b90b1127e');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-9447d626-5aec-4c92-964b-d4a1840c45f3\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-9447d626-5aec-4c92-964b-d4a1840c45f3')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-9447d626-5aec-4c92-964b-d4a1840c45f3 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "decision_tree_dataset"
            }
          },
          "metadata": {},
          "execution_count": 21
        }
      ],
      "source": [
        "X_train_dt = decision_tree_dataset[:int(len(decision_tree_dataset) * TRAIN_TEST_SPLIT_PERCENTAGE)]\n",
        "decision_tree_test = decision_tree_dataset[int(len(decision_tree_dataset) * TRAIN_TEST_SPLIT_PERCENTAGE):]\n",
        "\n",
        "\n",
        "X_test_dt = decision_tree_test.loc[:, (decision_tree_test.columns != 'NObesity')]\n",
        "Y_test_dt = decision_tree_test[\"NObesity\"]\n",
        "\n",
        "\n",
        "decision_tree_dataset"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Mg7iO7BrUtaF"
      },
      "outputs": [],
      "source": [
        "import math\n",
        "\n",
        "\n",
        "class DecisionTree:\n",
        "\n",
        "\n",
        "  # Classe per rappresentare i vari nodi di un albero\n",
        "  class TreeBranch:\n",
        "    def __init__(self, split_attribute_value, subtree):\n",
        "      self.subtree = subtree\n",
        "      self.split_attribute_value = split_attribute_value\n",
        "\n",
        "  # Classe che modella i branch di un albero\n",
        "  # ogni Branch ha un attributo di split (es. Weight)\n",
        "  # un sottoalbero se non è una foglia e un valore (output) se è una foglia\n",
        "  class DTreeStructure:\n",
        "      def __init__(self, split_attribute=\"\", *, value=None):\n",
        "        self.split_attribute = split_attribute\n",
        "        self.value = value\n",
        "        self.branches=[]\n",
        "\n",
        "      def add_branch(self, branch):\n",
        "        self.branches.append(branch)\n",
        "\n",
        "      def is_leaf_node(self):\n",
        "        return self.value is not None\n",
        "\n",
        "  def __init__(self,max_depth=10000, target=\"NObesity\"):\n",
        "    self.max_depth = max_depth\n",
        "    self.target = target\n",
        "    self.tree = []\n",
        "\n",
        "\n",
        "  def fit(self, X):\n",
        "    self.X_train = X\n",
        "    self.n_features = X.shape[1]\n",
        "    self.n_examples = X.shape[0]\n",
        "\n",
        "    # per evitare che nella lista di attributi su cui fare un possibile split\n",
        "    # sia presente l'attributo target\n",
        "    if(self.target in X.columns):\n",
        "      self.attributes = set(X.columns.drop(self.target))\n",
        "    else:\n",
        "      self.attributes = set(X.columns)\n",
        "\n",
        "    # funzione ricorsiva che crea il vero Decision Tree\n",
        "    self.tree = self._grow_tree(X, self.attributes, None)\n",
        "\n",
        "\n",
        "\n",
        "  def _grow_tree(self, examples, attributes, parent_examples, depth=0):\n",
        "\n",
        "    # Non ci sono esempi con questa combinazione di attributi -> si prende il valore più comune dell'albero padre\n",
        "    if examples.empty:\n",
        "      return self._most_common_label(parent_examples)\n",
        "\n",
        "    # Non ci sono più attributi su cui splittare oppure abbiamo raggiunto la profondità massima ->  si prende il valore più comune\n",
        "    if(not attributes or (depth>=self.max_depth)):\n",
        "      return self._most_common_label(examples)\n",
        "\n",
        "    # Si hanno soltanto esempi con la stessa classificazione -> si prende il valore più comune\n",
        "    if(len(examples[self.target].unique())==1):\n",
        "      v=examples[self.target].iloc[0]\n",
        "      return self.DTreeStructure(value=v)\n",
        "\n",
        "    # si prende l'attributo che ha più importanza\n",
        "    split_attribute = self._importance(examples, attributes)\n",
        "\n",
        "    # non si è trovato l'attributo, non si può splittare -> si prende il valore più comune\n",
        "    if split_attribute=='':\n",
        "      return self._most_common_label(examples)\n",
        "\n",
        "\n",
        "    #\n",
        "    tree = self.DTreeStructure(split_attribute)\n",
        "\n",
        "    split_attribute_values = examples[split_attribute].unique()\n",
        "\n",
        "    for value in split_attribute_values:\n",
        "\n",
        "      # Si prendono solo gli esempi che hanno lo stesso value per l'attributo split_attribute\n",
        "      exs = examples.loc[examples[split_attribute] == value]\n",
        "\n",
        "      # Si rimuove l'attributo scelto dal set degli attributi\n",
        "      attr = attributes - {split_attribute}\n",
        "\n",
        "      # Generare il sottoalbero ricorsivamente\n",
        "      subtree = self._grow_tree(exs, attr, examples, depth+1)\n",
        "\n",
        "      # Aggiungere il sottoalbero nella variabile branch del TreeBranch\n",
        "      tree.add_branch(self.TreeBranch(value, subtree))\n",
        "\n",
        "    return tree\n",
        "\n",
        "  def predict(self, X):\n",
        "    ret = []\n",
        "\n",
        "    # Per ogni riga di X calcola la predizione\n",
        "    for index,row in X.iterrows():\n",
        "      ret.append(self._predict_tree(row,self.tree))\n",
        "    return ret\n",
        "\n",
        "\n",
        "  def _predict_tree(self,sample,tree):\n",
        "\n",
        "    index = 0\n",
        "\n",
        "    # Si è arrivati alla foglia, viene returnata\n",
        "    if(tree.is_leaf_node()):\n",
        "      return tree.value\n",
        "\n",
        "\n",
        "    # Ciclo su tutti i sottoalberi di tree\n",
        "    for i in range(len(tree.branches)):\n",
        "      # Se il valore del sample per l'attributo split_attribute è uguale a split_attribute_value\n",
        "      # abbiamo trovato il sottoalbero giusto, si salva quindi l'indice\n",
        "      if tree.branches[i].split_attribute_value==sample.loc[tree.split_attribute]:\n",
        "        index = i\n",
        "        break\n",
        "\n",
        "    # si prende il sottoalbero corrispondente all'indice trovato prima\n",
        "    new_tree = tree.branches[index].subtree\n",
        "\n",
        "    # chiamata ricorsiva\n",
        "    return self._predict_tree(sample, new_tree)\n",
        "\n",
        "\n",
        "  # Information gain = Entropy - Remainder\n",
        "  def _information_gain(self, X, attribute):\n",
        "    information_gain = self._entropy(X)-self._remainder(X, attribute)\n",
        "    return information_gain\n",
        "\n",
        "  def _remainder(self,X,attribute):\n",
        "\n",
        "    remainder=0\n",
        "\n",
        "\n",
        "    # si prendono gli esempi del dataset con l'attributo attribute\n",
        "    groups=X.groupby(attribute)\n",
        "\n",
        "\n",
        "    for value,group in groups:\n",
        "\n",
        "      # per evitare divisione per 0\n",
        "      if(group.shape[0]!=0):\n",
        "\n",
        "        # Remainder = (dim_gruppo/dim_totale_dataset)*entropia(gruppo)\n",
        "        remainder+=(group.shape[0]/X.shape[0])*self._entropy(group)\n",
        "\n",
        "    return remainder\n",
        "\n",
        "\n",
        "  def _entropy(self, X):\n",
        "\n",
        "    # No entropia, informazione precisa\n",
        "    if (len(X[self.target].unique())==1):\n",
        "      return 0\n",
        "\n",
        "    # i valori sono solo 0,1 -> si usa la somma per trovare il numero di valori positivi\n",
        "    sum=X[self.target].sum()\n",
        "\n",
        "\n",
        "    probability=sum/(X.shape[0])\n",
        "\n",
        "    # formula entropia\n",
        "    entropy = -probability*math.log2(probability)-(1-probability)*math.log2(1-probability)\n",
        "    return entropy\n",
        "\n",
        "\n",
        "  def _importance(self, X, attributes):\n",
        "\n",
        "    max = 0\n",
        "    split_attribute = \"\"\n",
        "\n",
        "\n",
        "    # Ciclo su ogni attributo e calcolo quale di questi fornisce il maggior gain\n",
        "    # mediante funzione information_gain\n",
        "    for attr in attributes:\n",
        "      gain = self._information_gain(X, attr)\n",
        "      if(gain>max):\n",
        "        max = gain\n",
        "        split_attribute = attr\n",
        "    return split_attribute\n",
        "\n",
        "\n",
        "  def _most_common_label(self, X):\n",
        "\n",
        "    # Ritorna una foglia con il valore uguale al più comune in X\n",
        "    counter = Counter(X[self.target])\n",
        "    ret = counter.most_common(1)[0][0]\n",
        "    return self.DTreeStructure(value=ret)\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RKbyrDU1PVQj"
      },
      "source": [
        "## Neural Network\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZBY0N_S4rEEG"
      },
      "source": [
        "Nell'inizializzazione dei pesi vengono usate formule non viste a lezione (Inizializzazione He e inizializzazione Xavier). Questo perché effettuando alcuni test, generando valori totalmente randomici per i pesi, si ottenevano più frequentemente NaN in output, anche con piccoli valori di learning rate."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "DnKM5fcYPYOG"
      },
      "outputs": [],
      "source": [
        "class NeuralNetwork:\n",
        "\n",
        "  class Layer:\n",
        "\n",
        "    def __init__(self, activation, input_size, output_size):\n",
        "\n",
        "\n",
        "\n",
        "      if(activation==\"relu\"):\n",
        "        stddev = np.sqrt(2. / input_size)  # He initialization\n",
        "        self.weights = np.random.randn(input_size, output_size)*stddev\n",
        "      else:\n",
        "        stddev = np.sqrt(2. / (input_size + output_size)) # Xavier initialization\n",
        "        self.weights = np.random.randn(input_size, output_size)*stddev\n",
        "      self.bias = np.zeros(output_size)\n",
        "\n",
        "\n",
        "  def __init__(self, layers, activation=\"relu\", epoch=10000, alpha=0.01, l2=0.01, minibatch=0):\n",
        "\n",
        "    if(type(layers) is not list ):\n",
        "      raise Exception(\"Layer should be an array to indicate the structure of the Neural Network\")\n",
        "\n",
        "\n",
        "    if(activation==\"relu\"):\n",
        "      self.activation = self._relu\n",
        "    elif(activation==\"sigmoid\"):\n",
        "      self.activation = self._sigmoid\n",
        "    else:\n",
        "      raise Exception(\"Insert valid activation function [relu, sigmoid]\")\n",
        "\n",
        "    self.epoch=epoch\n",
        "    self.alpha=alpha\n",
        "    self.l2=l2\n",
        "    self.minibatch=minibatch\n",
        "\n",
        "\n",
        "    self.network = []\n",
        "    for i in range(len(layers)-1):\n",
        "        self.network.append(self.Layer(activation, layers[i], layers[i+1]))\n",
        "\n",
        "  def fit(self, X, y, print_mse=True):\n",
        "\n",
        "    # viene rimossa la colonna dummy bias\n",
        "    # perchè i bias vengono calcolati volta per volta nei layer\n",
        "\n",
        "    if(\"Bias\" in X.columns):\n",
        "      X_without_bias = X.drop(columns=[\"Bias\"])\n",
        "      X_t = np.array(X_without_bias)\n",
        "    else:\n",
        "      X_t = np.array(X)\n",
        "    y_t = np.array(y).reshape(len(y), 1)\n",
        "    x_number = X_t.shape[0]\n",
        "    predictions = np.random.randn(y_t.shape[0],1).reshape(-1)\n",
        "\n",
        "\n",
        "    for i in range(self.epoch):\n",
        "\n",
        "      # se minibatch = 0 il gradient descent viene effettuato su un minibatch del dataset iniziale\n",
        "      if(self.minibatch!=0):\n",
        "        if(print_mse and i%(self.epoch/10)==0):\n",
        "            print(\"Epoch: \"+str(i))\n",
        "\n",
        "        # creazione dataset randomico di dimensione = minibatch\n",
        "        permutation = np.random.permutation(x_number)\n",
        "        X_shuffled = X_t[permutation]\n",
        "        y_shuffled = y_t[permutation]\n",
        "\n",
        "        for j in range(0, x_number, self.minibatch):\n",
        "          end = j + self.minibatch\n",
        "          x_batch = X_shuffled[j:end]\n",
        "          y_batch = y_shuffled[j:end]\n",
        "\n",
        "          # #step di feedforward e backpropagation per trovare i migliori pesi e bias\n",
        "          outputs=self._feedforward(x_batch)\n",
        "          self._backpropagation(self.network, outputs, y_batch, self.minibatch)\n",
        "\n",
        "      else:\n",
        "          if(print_mse and i%(self.epoch/10)==0):\n",
        "            mse = mean_squared_error(denormalize(y_t.reshape(-1), r_train_min, r_train_max), denormalize(predictions, r_train_min, r_train_max))\n",
        "            print(\"Epoch: \"+str(i)+\" ---- MSE: \"+ str(mse))\n",
        "\n",
        "          #step di feedforward e backpropagation per trovare i migliori pesi e bias\n",
        "          outputs=self._feedforward(X_t)\n",
        "          predictions = outputs[-1].reshape(-1)\n",
        "          self._backpropagation(self.network, outputs, y_t, X_t.shape[0])\n",
        "      self.fitted = True\n",
        "\n",
        "  def predict(self, X):\n",
        "    if(\"Bias\" in X.columns):\n",
        "      X_t = np.array(X.drop(columns=[\"Bias\"]))\n",
        "    else:\n",
        "      X_t = np.array(X)\n",
        "    if(not self.fitted):\n",
        "      raise Exception(\"Fit the model first\")\n",
        "    return denormalize(self._feedforward(X_t)[-1].reshape(-1), r_train_min, r_train_max)\n",
        "\n",
        "\n",
        "  def _relu(self, x, derivative=False):\n",
        "\n",
        "    if(derivative):\n",
        "      return np.where(x > 0, 1, 0)\n",
        "\n",
        "    return np.maximum(0, x)\n",
        "\n",
        "\n",
        "  def _sigmoid(self, x, derivative=False):\n",
        "    if(derivative):\n",
        "      sig = self._sigmoid(x, False)\n",
        "      return sig*(1-sig)\n",
        "    return 1 / (1 + np.exp(-x))\n",
        "\n",
        "  def _feedforward(self, input):\n",
        "\n",
        "\n",
        "    # Per ogni layer (eccetto l'ultimo) viene calcolato f = X*W + b\n",
        "    # Viene passata f nella non linearità (sigmoide o relu)\n",
        "    # e fornita in input al prossimo layer\n",
        "\n",
        "    outputs = list([np.array(input)])\n",
        "    for i in range(len(self.network)-1):\n",
        "      weights = self.network[i].weights\n",
        "      bias = self.network[i].bias\n",
        "      output = self.activation(np.dot(outputs[i],weights)+bias)\n",
        "      outputs.append(output)\n",
        "\n",
        "\n",
        "\n",
        "    # L'ultimo layer calcola f = X*W + b ma non viene passata nella non linearità\n",
        "    output=np.dot(outputs[-1],self.network[-1].weights)+self.network[-1].bias\n",
        "    outputs.append(output)\n",
        "\n",
        "    return outputs\n",
        "\n",
        "\n",
        "  def _backpropagation(self, network,outputs,y, batch_size):\n",
        "\n",
        "\n",
        "    # Per l'algoritmo di backpropagation vengono calcolati i vari gradienti\n",
        "\n",
        "    # Sfrutta la chain rule:\n",
        "\n",
        "\n",
        "    # Si definiscono\n",
        "    # Theta = sigmoid / relu\n",
        "    # f_i = h_i * W_i + b_i\n",
        "    # h_0 = input, h_i = Theta(f_i)\n",
        "    #\n",
        "    # Si prende una variabile ausiliaria delta = dL/dfi\n",
        "    # delta = -2(y-pred) se i è un layer esterno, altrimenti delta = Theta'(fi)*dL/df_i+1 dot (W_i+1.T)\n",
        "\n",
        "    # dL/dWi = h_i.T dot delta\n",
        "    # dL/dBi = delta dot 1 dove 1 è il vettore unitario, quindi dL/dBi = somma righe di delta\n",
        "\n",
        "    for i in reversed(range(len(self.network))):\n",
        "      activated=outputs[i+1]\n",
        "      h_i=outputs[i]\n",
        "      if(i==(len(self.network)-1)):\n",
        "        delta = -(y-activated)\n",
        "      else:\n",
        "        activation_derivative = self.activation(activated, True)\n",
        "        delta = activation_derivative * np.dot(delta, self.network[i+1].weights.T)\n",
        "\n",
        "      gradient_weights =  (1/batch_size)*np.dot(h_i.T, delta)\n",
        "      gradient_biases = (1/batch_size)*np.sum(delta, axis=0)\n",
        "\n",
        "\n",
        "      network[i].bias = network[i].bias - self.alpha * gradient_biases\n",
        "      network[i].weights = network[i].weights - self.alpha*(gradient_weights + 2*self.l2 * network[i].weights)\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Zo3pfmGabY6f"
      },
      "source": [
        "##Hyperparameter Tuning\n",
        "\n",
        "Viene utilizzata una GridSearch che, presi in input una serie di iperparametri, ne calcola tutte le possibili combinazioni e allena il modello desiderato con ciascuna di queste combinazioni. Per i regressori sceglie gli iperparametri che forniscono il MSE più basso, mentre per i classificatori vengono scelti gli iperparametri che portano l'accuratezza maggiore. <br>\n",
        "\n",
        "Per la regressione, alcune combinazioni di parametri potrebbero far divergere il gradiente e quindi darebbero luogo a NaN in output. Per aggirare questo problema, viene usata una try-except che salta ai prossimi parametri se per quelli attuali è presente un errore."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ZN1pBH9vXZ4B"
      },
      "outputs": [],
      "source": [
        "import warnings\n",
        "## per ignorare i warning di numpy quando il gradiente diverge\n",
        "warnings.filterwarnings('ignore')\n",
        "\n",
        "import itertools\n",
        "\n",
        "def grid_search(model_type, X_train=[], y_train=[], X_test=[], y_test=[], params={}):\n",
        "    best_score = None\n",
        "    chosen_params = None\n",
        "    classification=False\n",
        "    # generare tutte le possibili combinazioni di iperparametri dal dizionario\n",
        "    h_params_comb = list(itertools.product(*params.values()))\n",
        "    for param_combination in h_params_comb:\n",
        "\n",
        "        # printa i parametri attuali\n",
        "        for i,key in enumerate(params.keys()):\n",
        "          print(key,': ', param_combination[i],end='; ',sep='')\n",
        "\n",
        "        # instanzia il modello cercato\n",
        "\n",
        "        # è tutto racchiuso in una try-except perché\n",
        "        # per alcuni parametri il gradiente potrebbe divergere e dar luogo a NaN in output\n",
        "        # quindi per evitare l'interruzione dell'esecuzione quando viene fuori un errore\n",
        "        # passo ai prossimi parametri\n",
        "        try:\n",
        "          if(model_type == \"linear_regressor\"):\n",
        "            model = LinearRegressor(*param_combination)\n",
        "            model.fit(X_train, y_train, False)\n",
        "          elif(model_type == \"logistic_classifier\"):\n",
        "            model = LogisticClassifier(*param_combination)\n",
        "            model.fit(X_train, y_train, False)\n",
        "            classification = True\n",
        "          elif(model_type == \"knn_classifier\"):\n",
        "            model = KNNClassifier(*param_combination)\n",
        "            model.fit(X_train, y_train)\n",
        "            classification = True\n",
        "          elif(model_type == \"decision_tree\"):\n",
        "            model = DecisionTree(*param_combination)\n",
        "            model.fit(X_train)\n",
        "            classification = True\n",
        "          elif(model_type == \"neural_network\"):\n",
        "            model = NeuralNetwork(*param_combination)\n",
        "            model.fit(X_train, y_train, False)\n",
        "          else:\n",
        "            raise Exception(\"Unknown model type\")\n",
        "            return\n",
        "          prediction = model.predict(X_test)\n",
        "          # stampare i parametri per questa iterazione\n",
        "\n",
        "\n",
        "          # printa lo score in base al tipo di modello (Classificatore/Regressore)\n",
        "          if(classification):\n",
        "            print('Accuracy:',end=' ')\n",
        "            score=accuracy_score(y_test,prediction)\n",
        "            print(score)\n",
        "            if best_score is None or score > best_score:\n",
        "              best_score = score\n",
        "              chosen_params = param_combination\n",
        "          else:\n",
        "            print('MSE:',end=' ')\n",
        "            score=mean_squared_error(prediction,y_test)\n",
        "            print(score)\n",
        "            # aggiornare il miglior punteggio e i migliori parametri se il punteggio corrente è migliore\n",
        "            if best_score is None or score < best_score:\n",
        "              best_score = score\n",
        "              chosen_params = param_combination\n",
        "\n",
        "        except Exception as e:\n",
        "          print(e)\n",
        "          continue\n",
        "\n",
        "    return chosen_params\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "o8Nrr1gshoS_"
      },
      "source": [
        "##Hyperparameter Tuning LinearRegression"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yJ-9okzd53pi"
      },
      "source": [
        "Per il LinearRegressor, gli iperparametri da scegliere sono tre: il numero di epoche, la learning rate, la dimensione del minibatch. Se minibatch = 0 si usa il Gradient Descent normale. Se epoch = 0, viene usata la Normal Equation che ovviamente fornisce il valore ottimale di MSE."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oY1eWAU7bXee",
        "outputId": "dd20748e-677d-4a1d-e4c0-16aa842bfa03"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 10000; Step: 0.001; Minibatch: 0; MSE: 3773.547073832318\n",
            "Epoch: 10000; Step: 0.001; Minibatch: 16; MSE: 39.25155116578878\n",
            "Epoch: 10000; Step: 0.001; Minibatch: 32; MSE: 39.18305047576322\n",
            "Epoch: 10000; Step: 0.001; Minibatch: 64; MSE: 41.59290263679296\n",
            "Epoch: 10000; Step: 0.01; Minibatch: 0; MSE: 78.25662105160096\n",
            "Epoch: 10000; Step: 0.01; Minibatch: 16; MSE: Input contains NaN.\n",
            "Epoch: 10000; Step: 0.01; Minibatch: 32; MSE: Input contains NaN.\n",
            "Epoch: 10000; Step: 0.01; Minibatch: 64; MSE: 39.41263256716261\n",
            "Epoch: 10000; Step: 0.1; Minibatch: 0; MSE: 39.28034139886174\n",
            "Epoch: 10000; Step: 0.1; Minibatch: 16; MSE: Input contains NaN.\n",
            "Epoch: 10000; Step: 0.1; Minibatch: 32; MSE: Input contains NaN.\n",
            "Epoch: 10000; Step: 0.1; Minibatch: 64; MSE: Input contains NaN.\n",
            "Epoch: 100000; Step: 0.001; Minibatch: 0; MSE: 86.08780171575428\n",
            "Epoch: 100000; Step: 0.001; Minibatch: 16; MSE: 39.35166135594208\n",
            "Epoch: 100000; Step: 0.001; Minibatch: 32; MSE: 39.29771896784832\n",
            "Epoch: 100000; Step: 0.001; Minibatch: 64; MSE: 43.259214953463996\n",
            "Epoch: 100000; Step: 0.01; Minibatch: 0; MSE: 39.267772978330946\n",
            "Epoch: 100000; Step: 0.01; Minibatch: 16; MSE: Input contains NaN.\n",
            "Epoch: 100000; Step: 0.01; Minibatch: 32; MSE: Input contains NaN.\n",
            "Epoch: 100000; Step: 0.01; Minibatch: 64; MSE: 39.11919600706575\n",
            "Epoch: 100000; Step: 0.1; Minibatch: 0; MSE: 39.28006316845677\n",
            "Epoch: 100000; Step: 0.1; Minibatch: 16; MSE: Input contains NaN.\n",
            "Epoch: 100000; Step: 0.1; Minibatch: 32; MSE: Input contains NaN.\n",
            "Epoch: 100000; Step: 0.1; Minibatch: 64; MSE: Input contains NaN.\n",
            "Epoch: 0 MSE: 325705.6768522973\n",
            "Epoch: 10000 MSE: 25.015653309007874\n",
            "Epoch: 20000 MSE: 25.015653309007874\n",
            "Epoch: 30000 MSE: 25.015653309007874\n",
            "Epoch: 40000 MSE: 25.015653309007874\n",
            "Epoch: 50000 MSE: 25.015653309007874\n",
            "Epoch: 60000 MSE: 25.015653309007874\n",
            "Epoch: 70000 MSE: 25.015653309007874\n",
            "Epoch: 80000 MSE: 25.015653309007874\n",
            "Epoch: 90000 MSE: 25.015653309007874\n",
            "MSE [NormalEquation]: 39.28006316845805\n",
            "MSE [GradientDescent]: 39.490711122383765\n"
          ]
        }
      ],
      "source": [
        "#epoch, step, minibatch\n",
        "epoch=[10**i for i in range(4, 6)]\n",
        "step = [10**i for i in range(-3, 0)]\n",
        "minibatch = [0,16,32,64]\n",
        "\n",
        "params = {\"Epoch\": epoch, \"Step\": step, \"Minibatch\": minibatch}\n",
        "\n",
        "chosen_params = grid_search(\"linear_regressor\", X_regression_train, Y_regression_train, X_regression_test, Y_regression_test, params)\n",
        "\n",
        "linear_regressor = LinearRegressor(*chosen_params)\n",
        "linear_regressor.fit(X_regression_train, Y_regression_train)\n",
        "\n",
        "#LinearRegressor con Normal Equation\n",
        "linear_regressor_normal_equation = LinearRegressor(epoch=0)\n",
        "linear_regressor_normal_equation.fit(X_regression_train, Y_regression_train)\n",
        "\n",
        "\n",
        "print(\"MSE [NormalEquation]: \"+str(mean_squared_error(Y_regression_test, linear_regressor_normal_equation.predict(X_regression_test))))\n",
        "print(\"MSE [GradientDescent]: \"+str(mean_squared_error(Y_regression_test, linear_regressor.predict(X_regression_test))))\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qLN14RgLhBRu"
      },
      "source": [
        "##Hyperparameter Tuning LogisticClassifier"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yD5ZmFJb7Bz7"
      },
      "source": [
        "Il LogisticClassifier ha due iperparametri: il numero di epoche, il valore della learning rate."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "O7RMyjbCcK4_",
        "outputId": "44c0d882-ebe5-4f80-e393-0e7f33f2d332"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 10000; Step: 0.0001; Accuracy: 0.5283018867924528\n",
            "Epoch: 10000; Step: 0.001; Accuracy: 0.7641509433962265\n",
            "Epoch: 10000; Step: 0.01; Accuracy: 0.8537735849056604\n",
            "Epoch: 10000; Step: 0.1; Accuracy: 0.9716981132075472\n",
            "Epoch: 100000; Step: 0.0001; Accuracy: 0.7452830188679245\n",
            "Epoch: 100000; Step: 0.001; Accuracy: 0.8773584905660378\n",
            "Epoch: 100000; Step: 0.01; Accuracy: 0.9716981132075472\n",
            "Epoch: 100000; Step: 0.1; Accuracy: 1.0\n",
            "(100000, 0.1)\n",
            "Epoch: 0 Accuracy: 0.507635597682991\n",
            "Epoch: 10000 Accuracy: 0.9826224328593997\n",
            "Epoch: 20000 Accuracy: 0.9921011058451816\n",
            "Epoch: 30000 Accuracy: 0.9931542917324908\n",
            "Epoch: 40000 Accuracy: 0.9936808846761453\n",
            "Epoch: 50000 Accuracy: 0.9942074776197999\n",
            "Epoch: 60000 Accuracy: 0.9942074776197999\n",
            "Epoch: 70000 Accuracy: 0.995260663507109\n",
            "Epoch: 80000 Accuracy: 0.9957872564507636\n",
            "Epoch: 90000 Accuracy: 0.9957872564507636\n"
          ]
        }
      ],
      "source": [
        "#epoch, step\n",
        "epoch=[10**i for i in range(4, 6)]\n",
        "step = [10**i for i in range(-4, 0)]\n",
        "\n",
        "params = {\"Epoch\": epoch, \"Step\": step}\n",
        "\n",
        "chosen_params = grid_search(\"logistic_classifier\", X_classification_train, Y_classification_train, X_classification_test, Y_classification_test, params)\n",
        "print(chosen_params)\n",
        "logistic_classifier = LogisticClassifier(*chosen_params)\n",
        "logistic_classifier.fit(X_classification_train, Y_classification_train)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TwLqK6EuhfHf"
      },
      "source": [
        "##Hyperparameter Tuning KNNClassifier\n",
        "Per un classificatore K-Nearest-Neighbour sono due gli iperparametri migliori da trovare: il numero di vicini da prendere in considerazione per fornire un risultato (K), e la metrica usata per la distanza. Le due distanze più usate in questo ambito sono L1 (distanza Manhattan) e L2 (distanza Euclidea).\n",
        "\n",
        "Il classificatore nel nostro caso sembra lavorar meglio con metrica L1. La miglior accuratezza si ha con K=1 ma in genere questo porta ad overfitting."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eKEI4pEihfHg",
        "outputId": "9f902c09-fc0e-45fd-9b77-6e10ee06509b"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "K: 1; DistanceType: 1; Accuracy: 0.9811320754716981\n",
            "K: 1; DistanceType: 2; Accuracy: 0.9339622641509434\n",
            "K: 3; DistanceType: 1; Accuracy: 0.9622641509433962\n",
            "K: 3; DistanceType: 2; Accuracy: 0.9433962264150944\n",
            "K: 5; DistanceType: 1; Accuracy: 0.9528301886792453\n",
            "K: 5; DistanceType: 2; Accuracy: 0.9150943396226415\n",
            "K: 7; DistanceType: 1; Accuracy: 0.9481132075471698\n",
            "K: 7; DistanceType: 2; Accuracy: 0.8915094339622641\n",
            "K: 9; DistanceType: 1; Accuracy: 0.9481132075471698\n",
            "K: 9; DistanceType: 2; Accuracy: 0.8915094339622641\n",
            "(1, 1)\n"
          ]
        }
      ],
      "source": [
        "#k, distance_type\n",
        "\n",
        "k = [1,3,5,7, 9]\n",
        "distance_type=[1,2]\n",
        "params = {\"K\": k, \"DistanceType\": distance_type}\n",
        "chosen_params=grid_search(\"knn_classifier\", X_classification_train, Y_classification_train, X_classification_test, Y_classification_test, params)\n",
        "print(chosen_params)\n",
        "knn_classifier = KNNClassifier(*chosen_params)\n",
        "knn_classifier.fit(X_classification_train, Y_classification_train)\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QsAdgN2ZhsAF"
      },
      "source": [
        "##Hyperparameter Tuning Decision Tree\n",
        "Per un Decision Tree l'unico iperparametro da scegliere è la profondità massima. Per il nostro dataset l'albero di decisione che si viene a formare ha profondità minore di 5, in quanto anche per valori di profondità massima maggiori di 5 si ottiene la stessa accuracy.\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PUgQJyhehsAI",
        "outputId": "2afc5450-ded5-4e92-baca-e6c86d183476"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "MaxDepth: 1; Accuracy: 0.8915094339622641\n",
            "MaxDepth: 5; Accuracy: 0.9528301886792453\n",
            "MaxDepth: 10; Accuracy: 0.9528301886792453\n",
            "MaxDepth: 100; Accuracy: 0.9528301886792453\n",
            "MaxDepth: 10000; Accuracy: 0.9528301886792453\n",
            "MaxDepth: 100000; Accuracy: 0.9528301886792453\n",
            "(5,)\n"
          ]
        }
      ],
      "source": [
        "#max_depth\n",
        "\n",
        "max_depth = [1, 5, 10, 100,10000,100000]\n",
        "\n",
        "params = {\"MaxDepth\": max_depth}\n",
        "chosen_params = grid_search(\"decision_tree\", X_train_dt, [], X_test_dt, Y_test_dt, params)\n",
        "print(chosen_params)\n",
        "decision_tree = DecisionTree(*chosen_params, \"NObesity\")\n",
        "decision_tree.fit(X_train_dt)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3V1cY4Lxio8V"
      },
      "source": [
        "##Hyperparameter Tuning Neural Network\n",
        "Trovare la migliore combinazione di iperparametri per una Rete Neurale è un procedimento che richiede molto tempo.\n",
        "I parametri da provare sono molti e la GridSearch ne prova tutte le varie combinazioni: si crea un'enorme lista di parametri da testare, a questo va aggiunto anche il tempo di training di una Rete Neurale, che può avere tempi molto lunghi.\n",
        "\n",
        "Scelgo un piccolo set di iperparametri da testare per mostrare il funzionamento della GridSearch con Reti Neurali.\n",
        "\n",
        "\n",
        "**Osservazione:**\n",
        "*con queste combinazioni di iperparametri impiega 1 ora*"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lfP19Xh7io8W",
        "outputId": "82b22cea-09e7-420d-fc7e-7f3f9e98426a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.001; L2: 1e-05; Minibatch: 0; MSE: 1224.7773817857376\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.001; L2: 1e-05; Minibatch: 32; MSE: 117.39103445245506\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.001; L2: 0.0001; Minibatch: 0; MSE: 1593.7949008798032\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.001; L2: 0.0001; Minibatch: 32; MSE: 107.69774592948221\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.01; L2: 1e-05; Minibatch: 0; MSE: 326.1464922230837\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.01; L2: 1e-05; Minibatch: 32; MSE: 39.495222344194275\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.01; L2: 0.0001; Minibatch: 0; MSE: 599.6813536564939\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.01; L2: 0.0001; Minibatch: 32; MSE: 29.504623948436322\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.1; L2: 1e-05; Minibatch: 0; MSE: 57.859256655413475\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.1; L2: 1e-05; Minibatch: 32; MSE: 22.17613643088093\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.1; L2: 0.0001; Minibatch: 0; MSE: 102.34924555708264\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.1; L2: 0.0001; Minibatch: 32; MSE: 25.18878941111076\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.001; L2: 1e-05; Minibatch: 0; MSE: 542.2311597211851\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.001; L2: 1e-05; Minibatch: 32; MSE: 44.09868746735798\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.001; L2: 0.0001; Minibatch: 0; MSE: 547.243009426561\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.001; L2: 0.0001; Minibatch: 32; MSE: 33.92233003858108\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.01; L2: 1e-05; Minibatch: 0; MSE: 51.789928941117275\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.01; L2: 1e-05; Minibatch: 32; MSE: 25.843944890922188\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.01; L2: 0.0001; Minibatch: 0; MSE: 73.11876414119533\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.01; L2: 0.0001; Minibatch: 32; MSE: 23.78154482322949\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.1; L2: 1e-05; Minibatch: 0; MSE: 34.76899747696877\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.1; L2: 1e-05; Minibatch: 32; MSE: 18.974632462107184\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.1; L2: 0.0001; Minibatch: 0; MSE: 40.42313987488236\n",
            "Layers: [26, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.1; L2: 0.0001; Minibatch: 32; MSE: 19.11479238942398\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.001; L2: 1e-05; Minibatch: 0; MSE: 1162.3164388586529\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.001; L2: 1e-05; Minibatch: 32; MSE: 109.1297353779494\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.001; L2: 0.0001; Minibatch: 0; MSE: 960.9787624056077\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.001; L2: 0.0001; Minibatch: 32; MSE: 134.13058680826862\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.01; L2: 1e-05; Minibatch: 0; MSE: 588.0629811156629\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.01; L2: 1e-05; Minibatch: 32; MSE: 37.54350622434219\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.01; L2: 0.0001; Minibatch: 0; MSE: 542.7506642854909\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.01; L2: 0.0001; Minibatch: 32; MSE: 38.75483814507812\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.1; L2: 1e-05; Minibatch: 0; MSE: 115.41725856623438\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.1; L2: 1e-05; Minibatch: 32; MSE: 39.00720758103043\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.1; L2: 0.0001; Minibatch: 0; MSE: 88.90212083747572\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.1; L2: 0.0001; Minibatch: 32; MSE: 37.91476969850511\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.001; L2: 1e-05; Minibatch: 0; MSE: 492.9200457208264\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.001; L2: 1e-05; Minibatch: 32; MSE: 40.7297022363669\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.001; L2: 0.0001; Minibatch: 0; MSE: 485.2376129681556\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.001; L2: 0.0001; Minibatch: 32; MSE: 39.60303858735414\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.01; L2: 1e-05; Minibatch: 0; MSE: 99.99137574730543\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.01; L2: 1e-05; Minibatch: 32; MSE: 41.43202555725776\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.01; L2: 0.0001; Minibatch: 0; MSE: 83.48332622383656\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.01; L2: 0.0001; Minibatch: 32; MSE: 38.11887045892159\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.1; L2: 1e-05; Minibatch: 0; MSE: 37.801194225625295\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.1; L2: 1e-05; Minibatch: 32; MSE: 38.610093549462206\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.1; L2: 0.0001; Minibatch: 0; MSE: 37.57600440733635\n",
            "Layers: [26, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.1; L2: 0.0001; Minibatch: 32; MSE: 38.12945369386146\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.001; L2: 1e-05; Minibatch: 0; MSE: 539.6118211574642\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.001; L2: 1e-05; Minibatch: 32; MSE: 113.45632162327837\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.001; L2: 0.0001; Minibatch: 0; MSE: 1032.078368493812\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.001; L2: 0.0001; Minibatch: 32; MSE: 140.4145068497594\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.01; L2: 1e-05; Minibatch: 0; MSE: 222.63707502452195\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.01; L2: 1e-05; Minibatch: 32; MSE: 31.18345207565013\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.01; L2: 0.0001; Minibatch: 0; MSE: 252.58378384962532\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.01; L2: 0.0001; Minibatch: 32; MSE: 31.329451790387342\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.1; L2: 1e-05; Minibatch: 0; MSE: 66.0604982992382\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.1; L2: 1e-05; Minibatch: 32; MSE: 18.783507597635474\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.1; L2: 0.0001; Minibatch: 0; MSE: Input contains NaN.\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 1000; Alpha: 0.1; L2: 0.0001; Minibatch: 32; MSE: 21.598359646746903\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.001; L2: 1e-05; Minibatch: 0; MSE: 227.17647836033987\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.001; L2: 1e-05; Minibatch: 32; MSE: 36.472925823535384\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.001; L2: 0.0001; Minibatch: 0; MSE: 235.70598868346528\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.001; L2: 0.0001; Minibatch: 32; MSE: 41.52488347306914\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.01; L2: 1e-05; Minibatch: 0; MSE: 68.88088697644316\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.01; L2: 1e-05; Minibatch: 32; MSE: 17.11273700825562\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.01; L2: 0.0001; Minibatch: 0; MSE: 83.97646614927636\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.01; L2: 0.0001; Minibatch: 32; MSE: 21.00007695594774\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.1; L2: 1e-05; Minibatch: 0; MSE: 27.829485679082353\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.1; L2: 1e-05; Minibatch: 32; MSE: 17.35942266550872\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.1; L2: 0.0001; Minibatch: 0; MSE: 26.743421055093794\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: relu; Epoch: 10000; Alpha: 0.1; L2: 0.0001; Minibatch: 32; MSE: 18.720685334341052\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.001; L2: 1e-05; Minibatch: 0; MSE: 692.8390457121424\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.001; L2: 1e-05; Minibatch: 32; MSE: 669.8680016346752\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.001; L2: 0.0001; Minibatch: 0; MSE: 712.2378811680729\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.001; L2: 0.0001; Minibatch: 32; MSE: 661.2648997844237\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.01; L2: 1e-05; Minibatch: 0; MSE: 666.1705989285553\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.01; L2: 1e-05; Minibatch: 32; MSE: 486.3376496738049\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.01; L2: 0.0001; Minibatch: 0; MSE: 671.8398705357633\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.01; L2: 0.0001; Minibatch: 32; MSE: 121.13889711686613\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.1; L2: 1e-05; Minibatch: 0; MSE: 672.4434550569969\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.1; L2: 1e-05; Minibatch: 32; MSE: 43.349894978289385\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.1; L2: 0.0001; Minibatch: 0; MSE: 666.6237102574364\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 1000; Alpha: 0.1; L2: 0.0001; Minibatch: 32; MSE: 41.229932236979906\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.001; L2: 1e-05; Minibatch: 0; MSE: 683.4743313269674\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.001; L2: 1e-05; Minibatch: 32; MSE: 151.93978626519728\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.001; L2: 0.0001; Minibatch: 0; MSE: 670.1188874160719\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.001; L2: 0.0001; Minibatch: 32; MSE: 214.1192774268567\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.01; L2: 1e-05; Minibatch: 0; MSE: 630.6353962032472\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.01; L2: 1e-05; Minibatch: 32; MSE: 37.58377537253224\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.01; L2: 0.0001; Minibatch: 0; MSE: 671.1821707255884\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.01; L2: 0.0001; Minibatch: 32; MSE: 44.24303187247699\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.1; L2: 1e-05; Minibatch: 0; MSE: 69.3184124410519\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.1; L2: 1e-05; Minibatch: 32; MSE: 39.28777281102053\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.1; L2: 0.0001; Minibatch: 0; MSE: 57.13756979846154\n",
            "Layers: [26, 16, 16, 16, 1]; Activation: sigmoid; Epoch: 10000; Alpha: 0.1; L2: 0.0001; Minibatch: 32; MSE: 37.056505230202276\n",
            "Epoch: 0\n",
            "Epoch: 1000\n",
            "Epoch: 2000\n",
            "Epoch: 3000\n",
            "Epoch: 4000\n",
            "Epoch: 5000\n",
            "Epoch: 6000\n",
            "Epoch: 7000\n",
            "Epoch: 8000\n",
            "Epoch: 9000\n"
          ]
        }
      ],
      "source": [
        "\n",
        "# layers, activation, epoch, alpha, l2, minibatch\n",
        "\n",
        "layers=[[26,16,1], [26,16,16,16,1]]\n",
        "activation = [\"relu\", \"sigmoid\"]\n",
        "epoch = [10**i for i in range(3, 5)]\n",
        "alpha = [10**i for i in range(-3,0)]\n",
        "l2 = [10**i for i in range(-5, -3)]\n",
        "minibatch = [0,32]\n",
        "params = {\"Layers\": layers, \"Activation\": activation, \"Epoch\": epoch, \"Alpha\": alpha, \"L2\": l2,  \"Minibatch\": minibatch}\n",
        "chosen_params = grid_search(\"neural_network\", X_regression_train, Y_regression_train, X_regression_test, Y_regression_test, params)\n",
        "neural_network = NeuralNetwork(*chosen_params)\n",
        "neural_network.fit(X_regression_train, Y_regression_train)\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LxubP58pfuNP"
      },
      "source": [
        "##Analisi finale e conclusioni"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MsG0vw6ef8BF"
      },
      "source": [
        "\n",
        "\n",
        "## 1.   Regressione\n",
        "\n",
        "Abbiamo a disposizione due modelli diversi: Linear Regressor e Neural Network. <br>\n",
        "Il regressore lineare è un modello molto semplice, non richiede molto tempo per scegliere gli iperparametri migliori ed è abbastanza veloce per quanto riguarda il tempo di addestramento. Ha un errore di circa 39kg sul test set. <br>\n",
        "La rete neurale è il modello che tra i due performa meglio. È più versatile, in quanto può avere più layer ciascuno con il proprio numero di neuroni, possiede inoltre delle non linearità, le funzioni di attivazione dei layer, che permettono di modellare al meglio il problema.\n",
        "Nonostante tutti questi fattori positivi le reti neurali presentano un grosso svantaggio dal punto di vista temporale. Il processo di Hyperparameter Tuning e di addestramento di una rete neurale può durare molto tempo, specialmente se il numero di layer, il numero di neuroni per ogni layer e il numero di esempi in un dataset è molto alto. <br>\n",
        "Vanno decise da subito le priorità: rete neurale se la priorità è la performance, regressore lineare se la priorità è il tempo di addestramento."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MK6uE8E9kVHd"
      },
      "source": [
        "\n",
        "## 2. Classificazione\n",
        "\n",
        "Sono tre i modelli addestrati per la classificazione: LogisticClassifier, DecisionTree, KNNClassifier. <br>\n",
        "Hanno tutti un accuracy molto alta che supera il 91%, ma questo è strettamente correlato al nostro dataset e al tipo di classificazione da fare. <br>\n",
        "Il \"migliore\" tra i tre è il LogisticRegressor, con un accuratezza che raggiunge anche il 100% in base allo split iniziale del dataset. Probabilmente con un dataset di training e test più grande l'accuratezza sarebbe scesa. <br>\n",
        "Possiamo ottenere più informazioni andando a guardare la confusion matrix e altre metriche."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "fNhS5rIkZTlT"
      },
      "outputs": [],
      "source": [
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.metrics import ConfusionMatrixDisplay\n",
        "from sklearn.metrics import precision_score\n",
        "from sklearn.metrics import f1_score\n",
        "from sklearn.metrics import roc_curve\n",
        "from sklearn.metrics import RocCurveDisplay\n",
        "from sklearn.metrics import auc\n",
        "from matplotlib import pyplot as plt\n",
        "\n",
        "knn_predictions = knn_classifier.predict(X_classification_test)\n",
        "decision_tree_predictions = decision_tree.predict(X_test_dt)\n",
        "logistic_predictions = logistic_classifier.predict(X_classification_test)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fkLFtbt7bACi"
      },
      "source": [
        "\n",
        "\n",
        "# LogisticRegressor\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XEreEXyaX2mg"
      },
      "source": [
        " Guardando la confusion matrix del LogisticRegressor classifica correttamente tutti gli esempi. È un caso molto speciale e probabilmente correlato alla dimensione del dataset e al tipo di classificazione\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "hVp-mJe5_qIn",
        "outputId": "3a09ea9c-2274-4c88-f44b-8af02d4fa5dc"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "confmatrix = confusion_matrix(Y_classification_test, logistic_predictions)\n",
        "disp = ConfusionMatrixDisplay(confusion_matrix=confmatrix)\n",
        "disp.plot()\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gEpoD_0aZ2Um"
      },
      "source": [
        "Altre metriche:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RcdgpjbwbGrL",
        "outputId": "825bdc88-43c3-435f-fa05-56fc0aed0fd6"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuratezza: 1.0\n",
            "Precisione: 1.0\n",
            "F1 Score: 1.0\n"
          ]
        }
      ],
      "source": [
        "print(\"Accuratezza: \"+str(accuracy_score(Y_classification_test, logistic_predictions)))\n",
        "print(\"Precisione: \"+str(precision_score(Y_classification_test, logistic_predictions)))\n",
        "print(\"F1 Score: \"+str(f1_score(Y_classification_test, logistic_predictions)))\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BCI5ozLl1u27"
      },
      "source": [
        "Visualizziamo la ROC"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "hOFaLXrTwdvo",
        "outputId": "222ab087-77ea-4f7e-942d-062e5a67ae9a"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "fpr,tpr, threshold = roc_curve(Y_classification_test, logistic_predictions)\n",
        "disp = RocCurveDisplay(fpr=fpr, tpr=tpr)\n",
        "disp.plot()\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lfwFqi0P_u1g"
      },
      "source": [
        "# KNNClassifier\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Szwpxsb-YXXY"
      },
      "source": [
        "Il KNNClassifier ha un'accuratezza che oscilla tra 95% e 96%. <br>\n",
        "Visualizziamo la Confusion Matrix\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "w7w_sFi9_u1h",
        "outputId": "51bc9ffa-1260-4926-e220-26a97942a994"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfsAAAGwCAYAAACuFMx9AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAwO0lEQVR4nO3deXhU9dn/8c9JQhYgCQQlIRAgLLIJQUDTiFXRaHBBEFprH+wTELEqi4DI0goIClHcMBZBUUEsLrQqFar0R2NlKYiC4qMVo0CUsCSomIQEszBzfn8gY6dAzeScZJhz3q/rOtflnGXmHo25c9/f7/kewzRNUwAAwLHCgh0AAACoXyR7AAAcjmQPAIDDkewBAHA4kj0AAA5HsgcAwOFI9gAAOFxEsAOwwuv16sCBA4qNjZVhGMEOBwAQINM0deTIESUnJyssrP7qz8rKSlVXV1t+n8jISEVHR9sQUcMK6WR/4MABpaSkBDsMAIBFhYWFatOmTb28d2VlpVLbNVXRIY/l90pKSlJBQUHIJfyQTvaxsbGSpD9vbqsmTRmRgDM92Cst2CEA9eaYarRJb/p+n9eH6upqFR3y6Kvt7RUXW/dcUXbEq3Z9v1R1dTXJviGdaN03aRqmJhb+AwJnsgijUbBDAOrPDwu2N8RQbNNYQ01j6/45XoXucHFIJ3sAAGrLY3rlsfA0GI/ptS+YBkayBwC4glemvKp7trdybbDR+wYAwOGo7AEAruCVV1Ya8dauDi6SPQDAFTymKY9Z91a8lWuDjTY+AAAOR2UPAHAFN0/QI9kDAFzBK1MelyZ72vgAADgclT0AwBVo4wMA4HDMxgcAAI5FZQ8AcAXvD5uV60MVyR4A4Aoei7PxrVwbbCR7AIAreExZfOqdfbE0NMbsAQBwOCp7AIArMGYPAIDDeWXII8PS9aGKNj4AAA5HZQ8AcAWveXyzcn2oItkDAFzBY7GNb+XaYKONDwCAw1HZAwBcwc2VPckeAOAKXtOQ17QwG9/CtcFGGx8AAIcj2QMAXOFEG9/KFogNGzZo0KBBSk5OlmEYWrVqld9x0zQ1c+ZMtWrVSjExMcrMzNQXX3zhd87hw4c1fPhwxcXFqVmzZho1apTKy8sD/u4kewCAK3gUZnkLREVFhdLS0rRw4cJTHp8/f75yc3O1ePFibd26VU2aNFFWVpYqKyt95wwfPlz/+te/tG7dOq1Zs0YbNmzQrbfeGvB3Z8weAOAKpsUxe/OHa8vKyvz2R0VFKSoq6qTzr7rqKl111VWneS9TCxYs0D333KPBgwdLkpYvX67ExEStWrVKN954o3bu3Km1a9fq/fffV79+/SRJTzzxhK6++mo9/PDDSk5OrnXsVPYAAAQgJSVF8fHxvi0nJyfg9ygoKFBRUZEyMzN9++Lj45Wenq4tW7ZIkrZs2aJmzZr5Er0kZWZmKiwsTFu3bg3o86jsAQCuYNetd4WFhYqLi/PtP1VV/1OKiookSYmJiX77ExMTfceKiorUsmVLv+MRERFKSEjwnVNbJHsAgCt4zDB5zLo3tE88zz4uLs4v2YcC2vgAADSwpKQkSVJxcbHf/uLiYt+xpKQkHTp0yO/4sWPHdPjwYd85tUWyBwC4gleGvAqzsNm3qE5qaqqSkpKUl5fn21dWVqatW7cqIyNDkpSRkaGSkhJt377dd87bb78tr9er9PT0gD6PNj4AwBUaernc8vJy7dq1y/e6oKBAO3bsUEJCgtq2basJEybo/vvvV+fOnZWamqoZM2YoOTlZQ4YMkSR169ZNAwcO1OjRo7V48WLV1NRo7NixuvHGGwOaiS+R7AEAqBfbtm3TgAEDfK8nTZokScrOztayZcs0ZcoUVVRU6NZbb1VJSYkuuugirV27VtHR0b5rVqxYobFjx+ryyy9XWFiYhg0bptzc3IBjMUzTDNkn9JaVlSk+Pl5v/V97NYllRALONKdDn2CHANSbY2aN3tFfVFpaWm+T3k7kitc/6qwmseF1fp+KIx5dn/ZFvcZaX6jsAQCucHzM3sKDcEL4qXeUwwAAOByVPQDAFbx1WN/e//qQHfUm2QMA3MH6ojokewAAzmgn7pev+/Whm+wZswcAwOGo7AEAruAxDXksPOLWyrXBRrIHALiCx+IEPQ9tfAAAcKaisgcAuILXDJPXwmx8L7PxAQA4s9HGBwAAjkVlDwBwBa+szaj32hdKgyPZAwBcwfqiOqHbDA/dyAEAQK1Q2QMAXMH62vihWx+T7AEAruDm59mT7AEAruDmyj50IwcAALVCZQ8AcAXri+qEbn1MsgcAuILXNOS1cp99CD/1LnT/TAEAALVCZQ8AcAWvxTZ+KC+qQ7IHALiC9afehW6yD93IAQBArVDZAwBcwSNDHgsL41i5NthI9gAAV6CNDwAAHIvKHgDgCh5Za8V77AulwZHsAQCu4OY2PskeAOAKPAgHAAA4FpU9AMAVTIvPsze59Q4AgDMbbXwAAOBYVPYAAFdw8yNuSfYAAFfwWHzqnZVrgy10IwcAALVCZQ8AcAXa+AAAOJxXYfJaaGhbuTbYQjdyAABQK1T2AABX8JiGPBZa8VauDTaSPQDAFRizBwDA4UyLT70zWUEPAACcqajsAQCu4JEhj4WH2Vi5NthI9gAAV/Ca1sbdvaaNwTQw2vgAADgclT1Oqao8TO88mqzP/l+8Kr5tpKQeR5U1Y59apx2VJJV/HaG8+a21e2OsKssi1O6CIxo4a59apFYFOXIgcL8aW6z+V5cqpVOVqivD9Om2xnp2bivt2x0d7NBgI6/FCXpWrg220I0c9Wr19Hba889YDXn0K9321k51uOiI/vibzioraiTTlF65rYO+2xupXz21R7eu2an41tX64286qfooP1IIPb0yKrR62VmacG1nTb+xg8IjTM17aY+iYjzBDg028sqwvIWqM+I388KFC9W+fXtFR0crPT1d7733XrBDcrWaSkM71zbT5VP3q90F5UpoX6VLJxxUQvsqbVtxlg4XRGn/h0119X2Fap12VGd1qNI19xWqpipMn6xuHuzwgYD9fngHrVuZoK8+j9aeT2P0yIS2SmxTo869vg92aIAtgp7sX3nlFU2aNEmzZs3SBx98oLS0NGVlZenQoUPBDs21vMcMmR5DEVH+s1Eiorwq3NZUx6oN3+sTjDApItJU4bamDRorUB+axB2v6I+UhAc5EtjpxAp6VrZQFfRk/+ijj2r06NEaOXKkunfvrsWLF6tx48Z67rnngh2aa0U19apNn3Jt/EOSjhQ3ktcj/d+qBO37sInKDzXSWR0rFZ9cpbcfaq3vS8PlqTb0z8WJKjsYqSOHGgU7fMASwzB12+z9+uS9xvoqPybY4cBGJ8bsrWyhKqgT9Kqrq7V9+3ZNnz7dty8sLEyZmZnasmXLSedXVVWpqurHCWBlZWUNEqcbDXnkS70xtZ0ey+gpI9xUqx5Hde6g73Twk8YKbyT9ctEerZ7WTg+dlyYj3FSH/mXqdEmpQvjOFECSNHbefrXrWqm7hnQKdiiAbYKa7L/55ht5PB4lJib67U9MTNRnn3120vk5OTmaPXt2Q4XnagntqjXi5S9UfTRMVeVhim15TH8el6pmKcf/2Eru+b1++9fPVFkWJk9NmJq0OKZnru+i5J5Hgxw5UHdj5u5T+hVluuv6jvrmYGSww4HNvLK4Nj4T9BrG9OnTVVpa6tsKCwuDHZLjRTb2KrblMX1fGq7dG2LV5YoSv+PRcV41aXFM3xZE6eDHjU86DoQGU2Pm7tOFA0s15ZcdVVwYFeyAUA9MizPxzRBO9kGt7M866yyFh4eruLjYb39xcbGSkpJOOj8qKkpRUfxP2BB2bYiVTEMtOlTq8JdR+vsDrXVWxyr1/sW3kqRP32ymxgnHFJ9crUP5MVo7p426XFGijj8/EuTIgcCNnbdfA67/TveOTNX35WFqfnaNJKniSLiqK0OqJsJ/wVPvgiQyMlJ9+/ZVXl6ehgwZIknyer3Ky8vT2LFjgxma61UdCdfbD7VWWVEjxcR71G3gdxpw1wGF/zD/7sihRvp/c9uo/JsIxZ5do15DD+visUXBDRqoo0Ejjv8R+/Bru/32PzwhRetWJgQjJMBWQV9Bb9KkScrOzla/fv10wQUXaMGCBaqoqNDIkSODHZqr9bimRD2uKTnt8fQRXyt9xNcNFxBQj7KS04IdAhpAQ6+g5/F4dO+99+qPf/yjioqKlJycrBEjRuiee+6RYRzvEpimqVmzZmnJkiUqKSlR//79tWjRInXu3LnOcZ5K0JP9r371K3399deaOXOmioqK1Lt3b61du/akSXsAAFjR0G38Bx98UIsWLdLzzz+vHj16aNu2bRo5cqTi4+M1fvx4SdL8+fOVm5ur559/XqmpqZoxY4aysrL06aefKjravuWag57sJWns2LG07QEAIeE/b/s+3XyyzZs3a/DgwbrmmmskSe3bt9dLL73kWyXWNE0tWLBA99xzjwYPHixJWr58uRITE7Vq1SrdeOONtsXMzBMAgCvYtTZ+SkqK4uPjfVtOTs4pP+/CCy9UXl6ePv/8c0nSRx99pE2bNumqq66SJBUUFKioqEiZmZm+a+Lj45Wenn7KtWasOCMqewAA6ptdbfzCwkLFxcX59p/uLrFp06aprKxMXbt2VXh4uDwej+bOnavhw4dLkoqKjk9qPtVaMyeO2YVkDwBAAOLi4vyS/emsXLlSK1as0IsvvqgePXpox44dmjBhgpKTk5Wdnd0Akf6IZA8AcIWGnqB39913a9q0ab6x9549e+qrr75STk6OsrOzfevJFBcXq1WrVr7riouL1bt37zrHeSqM2QMAXOFEsreyBeLo0aMKC/NPs+Hh4fJ6jz8xNDU1VUlJScrLy/MdLysr09atW5WRkWH9C/8bKnsAAOrBoEGDNHfuXLVt21Y9evTQhx9+qEcffVQ333yzJMkwDE2YMEH333+/Onfu7Lv1Ljk52bfQnF1I9gAAV2joNv4TTzyhGTNm6I477tChQ4eUnJys3/72t5o5c6bvnClTpqiiokK33nqrSkpKdNFFF2nt2rW23mMvSYZpmiH7VNKysjLFx8frrf9rryaxjEjAmeZ06BPsEIB6c8ys0Tv6i0pLS2s16a0uTuSKzDd/q4gmdX++yrGKKv396qfqNdb6QmUPAHAFNz8Ih3IYAACHo7IHALiCmyt7kj0AwBXcnOxp4wMA4HBU9gAAV3BzZU+yBwC4gmkaMi0kbCvXBhttfAAAHI7KHgDgCv/+TPq6Xh+qSPYAAFdw85g9bXwAAByOyh4A4ApunqBHsgcAuIKb2/gkewCAK7i5smfMHgAAh6OyBwC4gmmxjR/KlT3JHgDgCqYk07R2faiijQ8AgMNR2QMAXMErQwYr6AEA4FzMxgcAAI5FZQ8AcAWvachgUR0AAJzLNC3Oxg/h6fi08QEAcDgqewCAK7h5gh7JHgDgCiR7AAAczs0T9BizBwDA4ajsAQCu4ObZ+CR7AIArHE/2VsbsbQymgdHGBwDA4ajsAQCuwGx8AAAczpS1Z9KHcBefNj4AAE5HZQ8AcAXa+AAAOJ2L+/gkewCAO1is7BXClT1j9gAAOByVPQDAFVhBDwAAh3PzBD3a+AAAOByVPQDAHUzD2iS7EK7sSfYAAFdw85g9bXwAAByOyh4A4A4sqgMAgLO5eTZ+rZL9G2+8Ues3vO666+ocDAAAsF+tkv2QIUNq9WaGYcjj8ViJBwCA+hPCrXgrapXsvV5vfccBAEC9cnMb39Js/MrKSrviAACgfpk2bCEq4GTv8Xh03333qXXr1mratKn27NkjSZoxY4aeffZZ2wMEAADWBJzs586dq2XLlmn+/PmKjIz07T/33HP1zDPP2BocAAD2MWzYQlPAyX758uV6+umnNXz4cIWHh/v2p6Wl6bPPPrM1OAAAbEMbv/b279+vTp06nbTf6/WqpqbGlqAAAIB9Ak723bt318aNG0/a/+c//1nnnXeeLUEBAGA7F1f2Aa+gN3PmTGVnZ2v//v3yer167bXXlJ+fr+XLl2vNmjX1ESMAANa5+Kl3AVf2gwcP1urVq/X3v/9dTZo00cyZM7Vz506tXr1aV1xxRX3ECAAALKjT2vg///nPtW7dOrtjAQCg3vCI2zrYtm2bXnjhBb3wwgvavn27nTEBAGC/IIzZ79+/XzfddJNatGihmJgY9ezZU9u2bfsxJNPUzJkz1apVK8XExCgzM1NffPGFhS95agFX9vv27dOvf/1r/fOf/1SzZs0kSSUlJbrwwgv18ssvq02bNnbHCABAyPnuu+/Uv39/DRgwQG+99ZbOPvtsffHFF2revLnvnPnz5ys3N1fPP/+8UlNTNWPGDGVlZenTTz9VdHS0bbEEXNnfcsstqqmp0c6dO3X48GEdPnxYO3fulNfr1S233GJbYAAA2OrEBD0rm6SysjK/raqq6pQf9+CDDyolJUVLly7VBRdcoNTUVF155ZXq2LHj8XBMUwsWLNA999yjwYMHq1evXlq+fLkOHDigVatW2frVA07269ev16JFi9SlSxffvi5duuiJJ57Qhg0bbA0OAAC7GKb1TZJSUlIUHx/v23Jyck75eW+88Yb69eunX/7yl2rZsqXOO+88LVmyxHe8oKBARUVFyszM9O2Lj49Xenq6tmzZYut3D7iNn5KScsrFczwej5KTk20JCgAA21m9V/6HawsLCxUXF+fbHRUVdcrT9+zZo0WLFmnSpEn63e9+p/fff1/jx49XZGSksrOzVVRUJElKTEz0uy4xMdF3zC4BV/YPPfSQxo0b5zfBYNu2bbrzzjv18MMP2xocAABnmri4OL/tdMne6/WqT58+mjdvns477zzdeuutGj16tBYvXtzAEdeysm/evLkM48fFBCoqKpSenq6IiOOXHzt2TBEREbr55ps1ZMiQegkUAABLGnhRnVatWql79+5++7p166ZXX31VkpSUlCRJKi4uVqtWrXznFBcXq3fv3nWP8xRqlewXLFhg64cCANDgbGrj11b//v2Vn5/vt+/zzz9Xu3btJEmpqalKSkpSXl6eL7mXlZVp69atuv322y0EerJaJfvs7GxbPxQAAKebOHGiLrzwQs2bN0833HCD3nvvPT399NN6+umnJUmGYWjChAm6//771blzZ9+td8nJybZ3yeu0gt4JlZWVqq6u9tv375MWAAA4YzRwZX/++efr9ddf1/Tp0zVnzhylpqZqwYIFGj58uO+cKVOmqKKiQrfeeqtKSkp00UUXae3atbbeYy/VIdlXVFRo6tSpWrlypb799tuTjns8HlsCAwDAVg2c7CXp2muv1bXXXnva44ZhaM6cOZozZ46FwH5awLPxp0yZorfffluLFi1SVFSUnnnmGc2ePVvJyclavnx5fcQIAAAsCLiyX716tZYvX65LL71UI0eO1M9//nN16tRJ7dq104oVK/zaEwAAnDF4xG3tHT58WB06dJB0fHz+8OHDkqSLLrqIFfQAAGcsu1bQC0UBJ/sOHTqooKBAktS1a1etXLlS0vGK/8SDcQAAwJkj4GQ/cuRIffTRR5KkadOmaeHChYqOjtbEiRN199132x4gAAC2CMIjbs8UAY/ZT5w40ffPmZmZ+uyzz7R9+3Z16tRJvXr1sjU4AABgnaX77CWpXbt2vtWAAAA4UxmyNu4eutPzapnsc3Nza/2G48ePr3MwAADAfrVK9o899lit3swwjKAk+wd7pSnCaNTgnws0hL8d2BHsEIB6U3bEq+bnNNCHufjWu1ol+xOz7wEACFlBWEHvTBHwbHwAABBaLE/QAwAgJLi4sifZAwBcweoqeK5aQQ8AAIQWKnsAgDu4uI1fp8p+48aNuummm5SRkaH9+/dLkl544QVt2rTJ1uAAALCNi5fLDTjZv/rqq8rKylJMTIw+/PBDVVVVSZJKS0s1b9482wMEAADWBJzs77//fi1evFhLlixRo0Y/LmTTv39/ffDBB7YGBwCAXdz8iNuAx+zz8/N18cUXn7Q/Pj5eJSUldsQEAID9XLyCXsCVfVJSknbt2nXS/k2bNqlDhw62BAUAgO0Ys6+90aNH684779TWrVtlGIYOHDigFStWaPLkybr99tvrI0YAAGBBwG38adOmyev16vLLL9fRo0d18cUXKyoqSpMnT9a4cePqI0YAACxz86I6ASd7wzD0+9//Xnfffbd27dql8vJyde/eXU2bNq2P+AAAsIeL77Ov86I6kZGR6t69u52xAACAehBwsh8wYIAM4/QzEt9++21LAQEAUC+s3j7npsq+d+/efq9ramq0Y8cOffLJJ8rOzrYrLgAA7EUbv/Yee+yxU+6/9957VV5ebjkgAABgL9ueenfTTTfpueees+vtAACwl4vvs7ftqXdbtmxRdHS0XW8HAICtuPUuAEOHDvV7bZqmDh48qG3btmnGjBm2BQYAAOwRcLKPj4/3ex0WFqYuXbpozpw5uvLKK20LDAAA2COgZO/xeDRy5Ej17NlTzZs3r6+YAACwn4tn4wc0QS88PFxXXnklT7cDAIQcNz/iNuDZ+Oeee6727NlTH7EAAIB6EHCyv//++zV58mStWbNGBw8eVFlZmd8GAMAZy4W33UkBjNnPmTNHd911l66++mpJ0nXXXee3bK5pmjIMQx6Px/4oAQCwysVj9rVO9rNnz9Ztt92mf/zjH/UZDwAAsFmtk71pHv+T5pJLLqm3YAAAqC8sqlNL/+1pdwAAnNFo49fOOeec85MJ//Dhw5YCAgAA9goo2c+ePfukFfQAAAgFtPFr6cYbb1TLli3rKxYAAOqPi9v4tb7PnvF6AABCU8Cz8QEACEkuruxrney9Xm99xgEAQL1izB4AAKdzcWUf8Nr4AAAgtFDZAwDcwcWVPckeAOAKbh6zp40PAIDDUdkDANyBNj4AAM5GGx8AADgWlT0AwB1o4wMA4HAuTva08QEAcDgqewCAKxg/bFauD1UkewCAO9DGBwDA2U7cemdlq6sHHnhAhmFowoQJvn2VlZUaM2aMWrRooaZNm2rYsGEqLi62/kVPgWQPAEA9ev/99/XUU0+pV69efvsnTpyo1atX609/+pPWr1+vAwcOaOjQofUSA8keAOAOpg2bpLKyMr+tqqrqtB9ZXl6u4cOHa8mSJWrevLlvf2lpqZ599lk9+uijuuyyy9S3b18tXbpUmzdv1rvvvmv3NyfZAwBcxGKil6SUlBTFx8f7tpycnNN+3JgxY3TNNdcoMzPTb//27dtVU1Pjt79r165q27attmzZYsMX9ccEPQAAAlBYWKi4uDjf66ioqFOe9/LLL+uDDz7Q+++/f9KxoqIiRUZGqlmzZn77ExMTVVRUZGu8EskeAOASdq2NHxcX55fsT6WwsFB33nmn1q1bp+jo6Lp/qE1o4wMA3MGmMfva2L59uw4dOqQ+ffooIiJCERERWr9+vXJzcxUREaHExERVV1erpKTE77ri4mIlJSVZ+56nQGUPAIDNLr/8cn388cd++0aOHKmuXbtq6tSpSklJUaNGjZSXl6dhw4ZJkvLz87V3715lZGTYHg/JHgDgCg35iNvY2Fide+65fvuaNGmiFi1a+PaPGjVKkyZNUkJCguLi4jRu3DhlZGToZz/7Wd2DPA2SPQDAHc6wFfQee+wxhYWFadiwYaqqqlJWVpaefPJJez/kByR7AAAawDvvvOP3Ojo6WgsXLtTChQvr/bNJ9gAAV2jINv6ZhmQPAHCHM6yN35BI9gAAd3Bxsuc+ewAAHI7KHgDgCozZAwDgdLTxAQCAU1HZAwBcwTBNGWbdy3Mr1wYbyR4A4A608QEAgFNR2QMAXIHZ+AAAOB1tfAAA4FRU9gAAV6CNDwCA07m4jU+yBwC4gpsre8bsAQBwOCp7AIA70MYHAMD5QrkVbwVtfAAAHI7KHgDgDqZ5fLNyfYgi2QMAXIHZ+AAAwLGo7AEA7sBsfAAAnM3wHt+sXB+qaOMDAOBwVPb4Sb8aW6z+V5cqpVOVqivD9Om2xnp2bivt2x0d7NCAWvn43Sb605Mt9cXHjXW4uJFmPVugC68q9R03TWn5Q0la+2ILlZeFq3u/Co1/oFCtO1T7zvni/2L07Nxkff5RY4WFm7ro6hL99t4DimkSwuWe27i4jU9lj5/UK6NCq5edpQnXdtb0GzsoPMLUvJf2KCrGE+zQgFqpPBqmDj2+19h5+055fOXClvrLc2dr3AOFenzN54pu7NXv/qejqisNSdK3RRGadmNHJadW6fE1n2vuit36Kj9aD09o25BfAxadmI1vZQtVQU32GzZs0KBBg5ScnCzDMLRq1apghoPT+P3wDlq3MkFffR6tPZ/G6JEJbZXYpkade30f7NCAWjn/siMaMbVI/f+tmj/BNKVVz5ytX99ZpAsHlqlD90pNyf1K3xY30ua18ZKkrX+PV0SEqbHz9imlU5W69P5e4x/cp01/bab9BZEN/XVQVyfus7eyhaigJvuKigqlpaVp4cKFwQwDAWoSd7yiP1ISHuRIAOuK9kbq8KFG6vPzct++JnFedT3vqHZubyJJqqkyFNHIVNi//caMjD7evv/Xe00bNF6gLoI6Zn/VVVfpqquuqvX5VVVVqqqq8r0uKyurj7DwXxiGqdtm79cn7zXWV/kxwQ4HsOzwoeO/BpudXeO3v9nZNb5jaReV66nZrfWnJ8/WkFu+UeXRMD03L9nvepz5WFQnROTk5Cg+Pt63paSkBDsk1xk7b7/ada1Uzu3tgh0K0GDad6nU5AVf6dWnWuq6jr306949lJRSreZn18gwgh0das20YQtRIfUn6fTp0zVp0iTf67KyMhJ+Axozd5/SryjTXdd31DcHGaeEMyS0PCZJKvm6kVokHvPtL/m6kTr2+HFeymVDS3TZ0BJ993WEoht7ZRjSa0+frVbtqk56T+BME1LJPioqSlFRUcEOw4VMjZm7XxcOLNXdv+ik4kL+G8A5ktpWK6FljT7c1FQdzz2e3CuOhOmzDxvr2v/95qTzm599/A+Cv72UoEZRXvW5uPykc3BmcnMbP6SSPYJj7Lz9GnD9d7p3ZKq+Lw9T8x/GNiuOhKu6MqRGguBS31eE6UDBj3+kFhVGavcnMYptdkwt29RoyC1f66XHE9U6tUpJbav1/PxWapFYowsH/jh7/y/PnaXu/SoU08SrDzbE6pn7knXz7w6oaTy3oIYMnnoHnN6gEd9Kkh5+bbff/ocnpGjdyoRghAQE5POPGmvKLzr5Xj91b2tJ0hU3HNbkBXt1w5hDqjwapsenpKi8LFw9zq/Q3BV7FBn94y/3/B2N9cIjSaqsCFObTlUaP79Qmb/4rsG/C1AXQU325eXl2rVrl+91QUGBduzYoYSEBLVty2IVZ4qs5LRghwBYknZhuf52YMdpjxuGlD2lSNlTik57zpTcvfUQGRoSbfwg2bZtmwYMGOB7fWLyXXZ2tpYtWxakqAAAjuTi5XKDmuwvvfRSmSE8BgIAQChgzB4A4Aq08QEAcDqveXyzcn2IItkDANzBxWP23CQNAIDDUdkDAFzBkMUxe9siaXgkewCAO7h4BT3a+AAAOByVPQDAFbj1DgAAp2M2PgAAcCoqewCAKximKcPCJDsr1wYbyR4A4A7eHzYr14co2vgAADgclT0AwBVo4wMA4HQuno1PsgcAuAMr6AEAAKeisgcAuAIr6AEA4HS08QEAgJ1ycnJ0/vnnKzY2Vi1bttSQIUOUn5/vd05lZaXGjBmjFi1aqGnTpho2bJiKi4ttj4VkDwBwBcNrfQvE+vXrNWbMGL377rtat26dampqdOWVV6qiosJ3zsSJE7V69Wr96U9/0vr163XgwAENHTrU5m9OGx8A4BYN3MZfu3at3+tly5apZcuW2r59uy6++GKVlpbq2Wef1YsvvqjLLrtMkrR06VJ169ZN7777rn72s5/VPdb/QGUPAEAAysrK/LaqqqpaXVdaWipJSkhIkCRt375dNTU1yszM9J3TtWtXtW3bVlu2bLE1ZpI9AMAdTBs2SSkpKYqPj/dtOTk5P/nRXq9XEyZMUP/+/XXuuedKkoqKihQZGalmzZr5nZuYmKiioiKr39YPbXwAgCvYtVxuYWGh4uLifPujoqJ+8toxY8bok08+0aZNm+r8+VaQ7AEACEBcXJxfsv8pY8eO1Zo1a7Rhwwa1adPGtz8pKUnV1dUqKSnxq+6Li4uVlJRkZ8i08QEALnFigp6VLaCPMzV27Fi9/vrrevvtt5Wamup3vG/fvmrUqJHy8vJ8+/Lz87V3715lZGTY8pVPoLIHALiDKWvPpA9wBGDMmDF68cUX9Ze//EWxsbG+cfj4+HjFxMQoPj5eo0aN0qRJk5SQkKC4uDiNGzdOGRkZts7El0j2AACXaOhH3C5atEiSdOmll/rtX7p0qUaMGCFJeuyxxxQWFqZhw4apqqpKWVlZevLJJ+sc4+mQ7AEAqAdmLf44iI6O1sKFC7Vw4cJ6jYVkDwBwB1MWF9WxLZIGR7IHALgDD8IBAABORWUPAHAHryTD4vUhimQPAHCFhp6NfyahjQ8AgMNR2QMA3MHFE/RI9gAAd3BxsqeNDwCAw1HZAwDcwcWVPckeAOAO3HoHAICzcesdAABwLCp7AIA7MGYPAIDDeU3JsJCwvaGb7GnjAwDgcFT2AAB3oI0PAIDTWUz2Ct1kTxsfAACHo7IHALgDbXwAABzOa8pSK57Z+AAA4ExFZQ8AcAfTe3yzcn2IItkDANyBMXsAAByOMXsAAOBUVPYAAHegjQ8AgMOZspjsbYukwdHGBwDA4ajsAQDuQBsfAACH83olWbhX3hu699nTxgcAwOGo7AEA7kAbHwAAh3NxsqeNDwCAw1HZAwDcwcXL5ZLsAQCuYJpemRaeXGfl2mAj2QMA3ME0rVXnjNkDAIAzFZU9AMAdTItj9iFc2ZPsAQDu4PVKhoVx9xAes6eNDwCAw1HZAwDcgTY+AADOZnq9Mi208UP51jva+AAAOByVPQDAHWjjAwDgcF5TMtyZ7GnjAwDgcFT2AAB3ME1JVu6zD93KnmQPAHAF02vKtNDGN0n2AACc4UyvrFX23HoHAADOUFT2AABXoI0PAIDTubiNH9LJ/sRfWcdUY2mdBOBMVnYkdH/BAD+lrPz4z3dDVM1Wc8Ux1dgXTAML6WR/5MgRSdImvRnkSID60/ycYEcA1L8jR44oPj6+Xt47MjJSSUlJ2lRkPVckJSUpMjLShqgalmGG8CCE1+vVgQMHFBsbK8Mwgh2OK5SVlSklJUWFhYWKi4sLdjiArfj5bnimaerIkSNKTk5WWFj9zRmvrKxUdXW15feJjIxUdHS0DRE1rJCu7MPCwtSmTZtgh+FKcXFx/DKEY/Hz3bDqq6L/d9HR0SGZpO3CrXcAADgcyR4AAIcj2SMgUVFRmjVrlqKiooIdCmA7fr7hVCE9QQ8AAPw0KnsAAByOZA8AgMOR7AEAcDiSPQAADkeyR60tXLhQ7du3V3R0tNLT0/Xee+8FOyTAFhs2bNCgQYOUnJwswzC0atWqYIcE2Ipkj1p55ZVXNGnSJM2aNUsffPCB0tLSlJWVpUOHDgU7NMCyiooKpaWlaeHChcEOBagX3HqHWklPT9f555+vP/zhD5KOP5cgJSVF48aN07Rp04IcHWAfwzD0+uuva8iQIcEOBbANlT1+UnV1tbZv367MzEzfvrCwMGVmZmrLli1BjAwAUBske/ykb775Rh6PR4mJiX77ExMTVVRUFKSoAAC1RbIHAMDhSPb4SWeddZbCw8NVXFzst7+4uFhJSUlBigoAUFske/ykyMhI9e3bV3l5eb59Xq9XeXl5ysjICGJkAIDaiAh2AAgNkyZNUnZ2tvr166cLLrhACxYsUEVFhUaOHBns0ADLysvLtWvXLt/rgoIC7dixQwkJCWrbtm0QIwPswa13qLU//OEPeuihh1RUVKTevXsrNzdX6enpwQ4LsOydd97RgAEDTtqfnZ2tZcuWNXxAgM1I9gAAOBxj9gAAOBzJHgAAhyPZAwDgcCR7AAAcjmQPAIDDkewBAHA4kj0AAA5HsgcAwOFI9oBFI0aM0JAhQ3yvL730Uk2YMKHB43jnnXdkGIZKSkpOe45hGFq1alWt3/Pee+9V7969LcX15ZdfyjAM7dixw9L7AKg7kj0cacSIETIMQ4ZhKDIyUp06ddKcOXN07Nixev/s1157Tffdd1+tzq1NggYAq3gQDhxr4MCBWrp0qaqqqvTmm29qzJgxatSokaZPn37SudXV1YqMjLTlcxMSEmx5HwCwC5U9HCsqKkpJSUlq166dbr/9dmVmZuqNN96Q9GPrfe7cuUpOTlaXLl0kSYWFhbrhhhvUrFkzJSQkaPDgwfryyy997+nxeDRp0iQ1a9ZMLVq00JQpU/Sfj5f4zzZ+VVWVpk6dqpSUFEVFRalTp0569tln9eWXX/oevtK8eXMZhqERI0ZIOv4I4ZycHKWmpiomJkZpaWn685//7Pc5b775ps455xzFxMRowIABfnHW1tSpU3XOOeeocePG6tChg2bMmKGampqTznvqqaeUkpKixo0b64YbblBpaanf8WeeeUbdunVTdHS0unbtqieffDLgWADUH5I9XCMmJkbV1dW+13l5ecrPz9e6deu0Zs0a1dTUKCsrS7Gxsdq4caP++c9/qmnTpho4cKDvukceeUTLli3Tc889p02bNunw4cN6/fXX/+vn/u///q9eeukl5ebmaufOnXrqqafUtGlTpaSk6NVXX5Uk5efn6+DBg3r88cclSTk5OVq+fLkWL16sf/3rX5o4caJuuukmrV+/XtLxP0qGDh2qQYMGaceOHbrllls0bdq0gP+dxMbGatmyZfr000/1+OOPa8mSJXrsscf8ztm1a5dWrlyp1atXa+3atfrwww91xx13+I6vWLFCM2fO1Ny5c7Vz507NmzdPM2bM0PPPPx9wPADqiQk4UHZ2tjl48GDTNE3T6/Wa69atM6OioszJkyf7jicmJppVVVW+a1544QWzS5cuptfr9e2rqqoyY2JizL/97W+maZpmq1atzPnz5/uO19TUmG3atPF9lmma5iWXXGLeeeedpmmaZn5+vinJXLdu3Snj/Mc//mFKMr/77jvfvsrKSrNx48bm5s2b/c4dNWqU+etf/9o0TdOcPn262b17d7/jU6dOPem9/pMk8/XXXz/t8Yceesjs27ev7/WsWbPM8PBwc9++fb59b731lhkWFmYePHjQNE3T7Nixo/niiy/6vc99991nZmRkmKZpmgUFBaYk88MPPzzt5wKoX4zZw7HWrFmjpk2bqqamRl6vV//zP/+je++913e8Z8+efuP0H330kXbt2qXY2Fi/96msrNTu3btVWlqqgwcPKj093XcsIiJC/fr1O6mVf8KOHTsUHh6uSy65pNZx79q1S0ePHtUVV1zht7+6ulrnnXeeJGnnzp1+cUhSRkZGrT/jhFdeeUW5ubnavXu3ysvLdezYMcXFxfmd07ZtW7Vu3drvc7xer/Lz8xUbG6vdu3dr1KhRGj16tO+cY8eOKT4+PuB4ANQPkj0ca8CAAVq0aJEiIyOVnJysiAj/H/cmTZr4vS4vL1ffvn21YsWKk97r7LPPrlMMMTExAV9TXl4uSfrrX//ql2Sl4/MQ7LJlyxYNHz5cs2fPVlZWluLj4/Xyyy/rkUceCTjWJUuWnPTHR3h4uG2xArCGZA/HatKkiTp16lTr8/v06aNXXnlFLVu2PKm6PaFVq1baunWrLr74YknHK9jt27erT58+pzy/Z8+e8nq9Wr9+vTIzM086fqKz4PF4fPu6d++uqKgo7d2797QdgW7duvkmG57w7rvv/vSX/DebN29Wu3bt9Pvf/96376uvvjrpvL179+rAgQNKTk72fU5YWJi6dOmixMREJScna8+ePRo+fHhAnw+g4TBBD/jB8OHDddZZZ2nw4MHauHGjCgoK9M4772j8+PHat2+fJOnOO+/UAw88oFWrVumzzz7THXfc8V/vkW/fvr2ys7N18803a9WqVb73XLlypSSpXbt2MgxDa9as0ddff63y8nLFxsZq8uTJmjhxop5//nnt3r1bH3zwgZ544gnfpLfbbrtNX3zxhe6++27l5+frxRdf1LJlywL6vp07d9bevXv18ssva/fu3crNzT3lZMPo6GhlZ2fro48+0saNGzV+/HjdcMMNSkpKkiTNnj1bOTk5ys3N1eeff66PP/5YS5cu1aOPPhpQPADqD8ke+EHjxo21YcMGtW3bVkOHDlW3bt00atQoVVZW+ir9u+66S7/5zW+UnZ2tjIwMxcbG6vrrr/+v77to0SL94he/0B133KGuXbtq9OjRqqiokCS1bt1as2fP1rRp05SYmKixY8dKku677z7NmDFDOTk56tatmwYOHKi//vWvSk1NlXR8HP3VV1/VqlWrlJaWpsWLF2vevHkBfd/rrrtOEydO1NixY9W7d29t3rxZM2bMOOm8Tp06aejQobr66qt15ZVXqlevXn631t1yyy165plntHTpUvXs2VOXXHKJli1b5osVQPAZ5ulmFgEAAEegsgcAwOFI9gAAOBzJHgAAhyPZAwDgcCR7AAAcjmQPAIDDkewBAHA4kj0AAA5HsgcAwOFI9gAAOBzJHgAAh/v/yp5f+4HKpiwAAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "confmatrix = confusion_matrix(Y_classification_test, knn_predictions)\n",
        "disp = ConfusionMatrixDisplay(confusion_matrix=confmatrix)\n",
        "disp.plot()\n",
        "plt.show()\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JO2k21O_Zd_T"
      },
      "source": [
        "Altre metriche:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "R2UQKP9KZdUG",
        "outputId": "991d4170-d6c3-42d8-ccd9-0861a1bdb440"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuratezza: 0.9811320754716981\n",
            "Precisione: 0.9819819819819819\n",
            "F1 Score: 0.9819819819819819\n"
          ]
        }
      ],
      "source": [
        "print(\"Accuratezza: \"+str(accuracy_score(Y_classification_test, knn_predictions)))\n",
        "print(\"Precisione: \"+str(precision_score(Y_classification_test, knn_predictions)))\n",
        "print(\"F1 Score: \"+str(f1_score(Y_classification_test, knn_predictions)))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JrsSxQpW1uDo"
      },
      "source": [
        "Visualizziamo la ROC"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "gofzi2iK1Cfm",
        "outputId": "92ac69cf-dd98-4e65-df8a-e22bad60d43c"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "fpr,tpr, threshold = roc_curve(Y_classification_test, knn_predictions)\n",
        "disp = RocCurveDisplay(fpr=fpr, tpr=tpr)\n",
        "disp.plot()\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kTWYLzqQ_wCo"
      },
      "source": [
        "\n",
        "# DecisionTree\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w8JVf3q2Yhcq"
      },
      "source": [
        "Il DecisionTree ha un'accuratezza che oscilla tra 92%-95%, in base allo split iniziale del dataset, è quello che performa \"peggio\". Ha più False Negatives, questo significa che tende a classificare meglio gli esempi positivi. <br>\n",
        "Utilizzando un Random Tree Forest probabilmente si sarebbero ottenuti risultati decisamente migliori."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "mCEfMFOH_wCp",
        "outputId": "41ff4880-fdd3-4240-89c0-5dfbbf37e6b0"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "confmatrix = confusion_matrix(Y_test_dt, decision_tree_predictions)\n",
        "disp = ConfusionMatrixDisplay(confusion_matrix=confmatrix)\n",
        "disp.plot()\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Lj792zoNZnpy"
      },
      "source": [
        "Altre metriche:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VJq9aPkIZoqo",
        "outputId": "9111100b-6a26-427c-cfca-ca74eeb6cdf8"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuratezza: 0.9528301886792453\n",
            "Precisione: 0.9809523809523809\n",
            "F1 Score: 0.9537037037037037\n"
          ]
        }
      ],
      "source": [
        "print(\"Accuratezza: \"+str(accuracy_score(Y_test_dt, decision_tree_predictions)))\n",
        "print(\"Precisione: \"+str(precision_score(Y_test_dt, decision_tree_predictions)))\n",
        "print(\"F1 Score: \"+str(f1_score(Y_test_dt, decision_tree_predictions)))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KEHyIRfE1p2u"
      },
      "source": [
        "Visualizziamo la ROC"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "r6Qi_9kd1HGZ",
        "outputId": "e1118dd4-27e1-4010-d3be-29c534e49de6"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjcAAAGwCAYAAABVdURTAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA7S0lEQVR4nO3de3hU1bnH8d9MLpMASYCTknCJDaAgFCQCwgmoHDAKahFqqygUIipWuUhJUQGBgHKrCmIVpYCIcLAgFpUKwhEUC0iLAvEGQrkVBBJM0YTrTDKzzh+YgUjADM7szQzfz/PMc5ydvWfe2bV937P2etdyGGOMAAAAIoTT7gAAAACCieIGAABEFIobAAAQUShuAABARKG4AQAAEYXiBgAARBSKGwAAEFGi7Q7Aaj6fTwcOHFBCQoIcDofd4QAAgEowxujIkSOqU6eOnM7zj81ccsXNgQMHlJaWZncYAADgAuzbt0/16tU77zmXXHGTkJAg6dTNSUxMtDkaAABQGcXFxUpLS/Pn8fO55IqbskdRiYmJFDcAAISZykwpYUIxAACIKBQ3AAAgolDcAACAiEJxAwAAIgrFDQAAiCgUNwAAIKJQ3AAAgIhCcQMAACIKxQ0AAIgoFDcAACCi2Frc/P3vf1fXrl1Vp04dORwOvfXWWz96zerVq9WyZUu5XC5dfvnlmjNnTsjjBAAA4cPW4ubYsWNq0aKFpk2bVqnzd+/erVtvvVUdO3ZUXl6efv/73+v+++/XihUrQhwpAAAIF7ZunHnzzTfr5ptvrvT506dPV/369TV58mRJUpMmTbR27Vo9++yz6ty5c6jCBAAAP8LnMzrmKdVRd6mMkepUj7ctlrDaFXz9+vXKysoqd6xz5876/e9/f85r3G633G63/31xcXGowgMAIOyUeH065i7VkZOnCpOj7lIdPVmqI9//36Pukh+8P/Xyn3/GsTJt69fUwt9l2vabwqq4yc/PV0pKSrljKSkpKi4u1okTJxQff3aVOHHiRI0dO9aqEAEAsIS71Huq6Dj5g2KjomLknMVJiU6W+IIaV7TTIYcjqB8ZeAz2fn3oDR8+XDk5Of73xcXFSktLszEiAMClyhijEyXes4qN06MgJafe/0hhcvRkqTze4BYlrminEuKiVc0VrWpx0UpwxXz/f0+9P3287H2Mqrmiy11TzRUtV7RTDpurm7AqblJTU1VQUFDuWEFBgRITEysctZEkl8sll8tlRXgAgAjl8xkd9VRUjJwa/fjhI5ofFienrjlVuPhMcGOrGht1RvERc6r4OKPY+GHxkVBBYVLVFa3Y6MhZHSasipvMzEwtW7as3LH33ntPmZn2PdcDAFy8QjGfJBicDn1fXMSUKzz8IyMVFSMVFCtVY6MV5bT5GdBFyNbi5ujRo9qxY4f//e7du5WXl6eaNWvqsssu0/Dhw7V//37NnTtXkvTggw/qhRde0KOPPqp7771X77//vl5//XUtXbrUrp8AAAgyY4zcpb4fjHpcPPNJEuJOP5ap8JFNJQqTKrFRtj+6iWS2FjeffPKJOnbs6H9fNjcmOztbc+bM0cGDB7V3717/3+vXr6+lS5dqyJAheu6551SvXj3NmjWLNnAAuAgEYz5J2aObEm9wn93ExThPFSNlhchZ80fCZz4JfpzDGBPkp38Xt+LiYiUlJamoqEiJiYl2hwMAtvOWrU8SgfNJEuJOzSeJiYqc+SSXqkDyd1jNuQEAnPZT55McOXnq78c83qDGxXwS2I3iBgAsFCnzScoKl4QKHunExzCfBPaiuAGASvgp80mOnDGKwnwSIPQobgBEtHCeT5J4xijKWcWKi/kkwLlQ3AC4KAU6n+RIueKkxH8+80mASw/FDYCgCcf5JGfNGfl+FIX5JED4orgBcEHzSY6cPKM4YT4JgIsIxQ0QxsJhPslZXTUVFSPMJwEQRBQ3gA0qO5+kuMLihPkkAHA+FDdAJV3M80liohynC5LzPbZhPgmASwDFDSJepeeTnKs4CYP5JHExUUGNCwDCGcUNLloX83ySahU+omE+CQBcDChuEHQ/Np/kyBkjJeEwnyThB8VJ1dhoOZlPAgAXLYobSGI+CQAgclDchLnzzSc5csYoyLn2vLlY5pP8cDTlzOLEFc18EgBA5VHc2IT5JAAAhAbFjYW+PeZRn9kbtOubo5bOJ0n0Fx4VFCPMJwEARBiKGwtt/Pe3+nx/UbljzCcBACC4KG4s5C49NdE2I626Xs5uzXwSAABCgOLGQh7vqUdRCXHR+q9qLpujAQAgMjHj00Lu71ukXdHcdgAAQoUsayGP91RxE0txAwBAyJBlLeT5fs5NLC3SAACEDFnWQmUTiplEDABA6FDcWKisuOGxFAAAoUOWtZCH4gYAgJAjy1rIXXqqFZxuKQAAQocsayFGbgAACD2yrIUobgAACD2yrIXolgIAIPQobizEyA0AAKFHlrVQ2QrFLhbxAwAgZMiyFvJ3S8Vw2wEACBWyrIXYfgEAgNAjy1qIOTcAAIQeWdZCdEsBABB6FDcWYuQGAIDQI8taiI0zAQAIPbKshU4/luK2AwAQKmRZC3m+bwVn5AYAgNAhy1qobBE/WsEBAAgdsqxFjDGnH0uxiB8AACFDlrVIqc/ImFP/7IqiFRwAgFChuLFIWRu4xJwbAABCiSxrETfFDQAAliDLWqRs5Cba6VCU02FzNAAARC6KG4uwOjEAANYg01rEzRo3AABYgkxrEVYnBgDAGmRai/gX8KO4AQAgpMi0FnGXsDoxAABWINNapGzkxhXNAn4AAIQSxY1F6JYCAMAaZFqL0C0FAIA1yLQW8dAtBQCAJci0FqG4AQDAGmRai7iZcwMAgCVsz7TTpk1Tenq64uLi1LZtW23YsOG850+dOlWNGzdWfHy80tLSNGTIEJ08edKiaC/c6ZEbuqUAAAglW4ubhQsXKicnR7m5udq0aZNatGihzp0769ChQxWe/9prr2nYsGHKzc3V1q1b9fLLL2vhwoUaMWKExZEHzr+IH+vcAAAQUrZm2ilTpqhfv37q27evmjZtqunTp6tKlSqaPXt2hed/9NFHat++vXr27Kn09HTddNNNuvvuu8872uN2u1VcXFzuZQd3Cd1SAABYwbZM6/F4tHHjRmVlZZ0OxulUVlaW1q9fX+E17dq108aNG/3FzK5du7Rs2TLdcsst5/yeiRMnKikpyf9KS0sL7g+pJLeXCcUAAFgh2q4vLiwslNfrVUpKSrnjKSkp+uqrryq8pmfPniosLNS1114rY4xKS0v14IMPnvex1PDhw5WTk+N/X1xcbEuBwyJ+AABYI6wy7erVqzVhwgS9+OKL2rRpkxYvXqylS5fqySefPOc1LpdLiYmJ5V52oFsKAABr2DZyk5ycrKioKBUUFJQ7XlBQoNTU1AqvGTVqlHr37q37779fktS8eXMdO3ZMDzzwgB5//HE5nRdv4UC3FAAA1rCtGoiNjVWrVq20atUq/zGfz6dVq1YpMzOzwmuOHz9+VgETFXWqWDDGhC7YIOCxFAAA1rBt5EaScnJylJ2drdatW6tNmzaaOnWqjh07pr59+0qS+vTpo7p162rixImSpK5du2rKlCm6+uqr1bZtW+3YsUOjRo1S165d/UXOxYq9pQAAsIatxU2PHj30zTffaPTo0crPz1dGRoaWL1/un2S8d+/eciM1I0eOlMPh0MiRI7V//3797Gc/U9euXTV+/Hi7fkKlsf0CAADWcJiL/XlOkBUXFyspKUlFRUWWTi7uNesfWrfjP3rurgx1y6hr2fcCABAJAsnfDCNYxF3CCsUAAFiBTGuRsu0XXDHccgAAQolMaxF/t9RFPvEZAIBwR3FjERbxAwDAGmRai9AtBQCANci0FmHkBgAAa5BpLcIifgAAWINMaxEeSwEAYA0yrQWMMf5WcEZuAAAILTKtBUp9RmXrQLtoBQcAIKQobixQNplYYhE/AABCjUxrAc8ZxQ3bLwAAEFpkWguUFTfRToecTofN0QAAENkobixQ1gZOpxQAAKFHtrWAhwX8AACwDNnWAqxODACAdci2FnD7F/CjDRwAgFCjuLEAj6UAALAO2dYC/tWJaQMHACDkyLYWcJd83y3FAn4AAIQc2dYCjNwAAGAdsq0FmHMDAIB1yLYWoFsKAADrUNxYwOMvbrjdAACEGtnWAjyWAgDAOmRbC7C3FAAA1iHbWoCRGwAArEO2tYCbVnAAACxDtrWAu+T7CcUs4gcAQMiRbS1wehE/WsEBAAg1ihsLMOcGAADrkG0t4GadGwAALEO2tYDn+1ZwRm4AAAi9n5RtT548Gaw4IhqPpQAAsE7A2dbn8+nJJ59U3bp1Va1aNe3atUuSNGrUKL388stBDzAS8FgKAADrBJxtx40bpzlz5uipp55SbGys/3izZs00a9asoAYXKdhbCgAA6wScbefOnasZM2aoV69eijqjtblFixb66quvghpcpPC3glPcAAAQcgFn2/379+vyyy8/67jP51NJSUlQgoo0/kX8olnnBgCAUAu4uGnatKnWrFlz1vE33nhDV199dVCCijSM3AAAYJ3oQC8YPXq0srOztX//fvl8Pi1evFjbtm3T3Llz9c4774QixrDn75ZibykAAEIu4GzbrVs3/e1vf9PKlStVtWpVjR49Wlu3btXf/vY33XjjjaGIMey5v1/nhr2lAAAIvYBHbiTpuuuu03vvvRfsWCKWm5EbAAAsE3C2bdCggf7zn/+cdfy7775TgwYNghJUpGERPwAArBNwtt2zZ4+8Xu9Zx91ut/bv3x+UoCKJMeaMRfzolgIAINQq/VhqyZIl/n9esWKFkpKS/O+9Xq9WrVql9PT0oAYXCUq8xv/PjNwAABB6lS5uunfvLklyOBzKzs4u97eYmBilp6dr8uTJQQ0uEpS1gUusUAwAgBUqXdz4fKeSdP369fXxxx8rOTk5ZEFFEnfJ6Ud4TCgGACD0Au6W2r17dyjiiFhlIzcxUQ45nQ6bowEAIPJdUCv4sWPH9OGHH2rv3r3yeDzl/vbwww8HJbBIwQJ+AABYK+DiZvPmzbrlllt0/PhxHTt2TDVr1lRhYaGqVKmiWrVqUdz8gL9TKoZOKQAArBDwcMKQIUPUtWtXffvtt4qPj9c//vEP/fvf/1arVq30zDPPhCLGsMbIDQAA1go44+bl5ekPf/iDnE6noqKi5Ha7lZaWpqeeekojRowIRYxhzc0CfgAAWCrgjBsTEyOn89RltWrV0t69eyVJSUlJ2rdvX3CjiwD+faUobgAAsETAc26uvvpqffzxx7riiivUoUMHjR49WoWFhZo3b56aNWsWihjDGlsvAABgrYAz7oQJE1S7dm1J0vjx41WjRg099NBD+uabb/TnP/856AGGO4obAACsFfDITevWrf3/XKtWLS1fvjyoAUWa0/tKUdwAAGCFoGXcTZs26Ze//GXA102bNk3p6emKi4tT27ZttWHDhvOe/91332nAgAGqXbu2XC6XGjVqpGXLll1o2CF3euSGVnAAAKwQUHGzYsUKDR06VCNGjNCuXbskSV999ZW6d++ua665xr9FQ2UtXLhQOTk5ys3N1aZNm9SiRQt17txZhw4dqvB8j8ejG2+8UXv27NEbb7yhbdu2aebMmapbt25A32ulshWKaQUHAMAalX4s9fLLL6tfv36qWbOmvv32W82aNUtTpkzRoEGD1KNHD33xxRdq0qRJQF8+ZcoU9evXT3379pUkTZ8+XUuXLtXs2bM1bNiws86fPXu2Dh8+rI8++kgxMTGS9KM7kbvdbrndbv/74uLigGL8qcr2lnLFUNwAAGCFSmfc5557Tn/84x9VWFio119/XYWFhXrxxRf1+eefa/r06QEXNh6PRxs3blRWVtbpYJxOZWVlaf369RVes2TJEmVmZmrAgAFKSUlRs2bNNGHCBHm93grPl6SJEycqKSnJ/0pLSwsozp+qbOTGxcgNAACWqHTG3blzp+644w5J0u23367o6Gg9/fTTqlev3gV9cWFhobxer1JSUsodT0lJUX5+foXX7Nq1S2+88Ya8Xq+WLVumUaNGafLkyRo3btw5v2f48OEqKiryv6xei4duKQAArFXpx1InTpxQlSpVJEkOh0Mul8vfEm4Vn8+nWrVqacaMGYqKilKrVq20f/9+Pf3008rNza3wGpfLJZfLZWmcZ6JbCgAAawXUCj5r1ixVq1ZNklRaWqo5c+YoOTm53DmV3TgzOTlZUVFRKigoKHe8oKBAqampFV5Tu3ZtxcTEKCrqdOdRkyZNlJ+fL4/Ho9jY2EB+jiUYuQEAwFqVLm4uu+wyzZw50/8+NTVV8+bNK3eOw+GodHETGxurVq1aadWqVerevbukUyMzq1at0sCBAyu8pn379nrttdfk8/n8W0Bs375dtWvXvigLG4m9pQAAsFqli5s9e/YE/ctzcnKUnZ2t1q1bq02bNpo6daqOHTvm757q06eP6tatq4kTJ0qSHnroIb3wwgsaPHiwBg0apH/961+aMGFCpQsqO5x+LMU6NwAAWCHgFYqDqUePHvrmm280evRo5efnKyMjQ8uXL/dPMt67d69/hEaS0tLStGLFCg0ZMkRXXXWV6tatq8GDB+uxxx6z6yf8KB5LAQBgLYcxxtgdhJWKi4uVlJSkoqIiJSYmhvz7Bv1ls/726QGN/mVT3Xtt/ZB/HwAAkSiQ/M1wQoixiB8AANYi44YY2y8AAGAtMm6IMecGAABrXVDG3blzp0aOHKm7777bv8nlu+++qy+//DKowUUCuqUAALBWwMXNhx9+qObNm+uf//ynFi9erKNHj0qSPv3003OuEnwp87BCMQAAlgo44w4bNkzjxo3Te++9V27hvE6dOukf//hHUIOLBDyWAgDAWgFn3M8//1y/+tWvzjpeq1YtFRYWBiWoSOIu/b5biuIGAABLBJxxq1evroMHD551fPPmzapbt25QgookjNwAAGCtgDPuXXfdpccee0z5+flyOBzy+Xxat26dhg4dqj59+oQixrDmbwWnuAEAwBIBZ9wJEyboyiuvVFpamo4ePaqmTZvq+uuvV7t27TRy5MhQxBjW3CWscwMAgJUC3lsqNjZWM2fO1KhRo/TFF1/o6NGjuvrqq3XFFVeEIr6w5/5+5MYVQys4AABWCLi4Wbt2ra699lpddtlluuyyy0IRU8Qwxpyec8PIDQAAlgg443bq1En169fXiBEjtGXLllDEFDHK5ttIzLkBAMAqAWfcAwcO6A9/+IM+/PBDNWvWTBkZGXr66af19ddfhyK+sFY2aiPRCg4AgFUCzrjJyckaOHCg1q1bp507d+qOO+7Qq6++qvT0dHXq1CkUMYatM4sbHksBAGCNn5Rx69evr2HDhmnSpElq3ry5Pvzww2DFFRHK9pWKiXLI6XTYHA0AAJeGCy5u1q1bp/79+6t27drq2bOnmjVrpqVLlwYztrDnYdNMAAAsF3C31PDhw7VgwQIdOHBAN954o5577jl169ZNVapUCUV8YY0F/AAAsF7Axc3f//53PfLII7rzzjuVnJwcipgiBgv4AQBgvYCLm3Xr1oUijojk8X6/aWYMxQ0AAFapVHGzZMkS3XzzzYqJidGSJUvOe+5tt90WlMAigZsF/AAAsFylipvu3bsrPz9ftWrVUvfu3c95nsPhkPf70QqcUdww5wYAAMtUqrjx+XwV/jPO73S3FMUNAABWCTjrzp07V263+6zjHo9Hc+fODUpQkcLDyA0AAJYLOOv27dtXRUVFZx0/cuSI+vbtG5SgIsXpx1KscwMAgFUCLm6MMXI4zl5t9+uvv1ZSUlJQgooUPJYCAMB6lW4Fv/rqq+VwOORwOHTDDTcoOvr0pV6vV7t371aXLl1CEmS48pSemlzNYykAAKxT6eKmrEsqLy9PnTt3VrVq1fx/i42NVXp6un79618HPcBwVvZYykUrOAAAlql0cZObmytJSk9PV48ePRQXFxeyoCKF/7EUi/gBAGCZgFcozs7ODkUcEcm/txQjNwAAWKZSxU3NmjW1fft2JScnq0aNGhVOKC5z+PDhoAUX7ljEDwAA61WquHn22WeVkJDg/+fzFTc47XS3FK3gAABYpVLFzZmPou65555QxRJxGLkBAMB6AWfdTZs26fPPP/e/f/vtt9W9e3eNGDFCHo8nqMGFOzet4AAAWC7grPu73/1O27dvlyTt2rVLPXr0UJUqVbRo0SI9+uijQQ8wnLGIHwAA1gs4627fvl0ZGRmSpEWLFqlDhw567bXXNGfOHP31r38Ndnxhjb2lAACw3gVtv1C2M/jKlSt1yy23SJLS0tJUWFgY3OjCnH/ODa3gAABYJuCs27p1a40bN07z5s3Thx9+qFtvvVWStHv3bqWkpAQ9wHB2ehE/uqUAALBKwMXN1KlTtWnTJg0cOFCPP/64Lr/8cknSG2+8oXbt2gU9wHDGIn4AAFgv4BWKr7rqqnLdUmWefvppRUUxQnGmsm4pJhQDAGCdgIubMhs3btTWrVslSU2bNlXLli2DFlSkoFsKAADrBVzcHDp0SD169NCHH36o6tWrS5K+++47dezYUQsWLNDPfvazYMcYtuiWAgDAegFn3UGDBuno0aP68ssvdfjwYR0+fFhffPGFiouL9fDDD4cixrBFcQMAgPUCHrlZvny5Vq5cqSZNmviPNW3aVNOmTdNNN90U1ODCnZu9pQAAsFzAQwo+n08xMTFnHY+JifGvf4NTGLkBAMB6AWfdTp06afDgwTpw4ID/2P79+zVkyBDdcMMNQQ0u3Lm9FDcAAFgt4Kz7wgsvqLi4WOnp6WrYsKEaNmyo+vXrq7i4WM8//3woYgxLxhi6pQAAsEHAc27S0tK0adMmrVq1yt8K3qRJE2VlZQU9uHBWtoCfxMgNAABWCqi4WbhwoZYsWSKPx6MbbrhBgwYNClVcYa9s1EZihWIAAKxU6eLmpZde0oABA3TFFVcoPj5eixcv1s6dO/X000+HMr6w5T6juOGxFAAA1ql01n3hhReUm5urbdu2KS8vT6+++qpefPHFUMYW1jxn7AjucDhsjgYAgEtHpYubXbt2KTs72/++Z8+eKi0t1cGDB0MSWLijDRwAAHtUOvO63W5VrVr19IVOp2JjY3XixImQBBbu3HRKAQBgi4AmFI8aNUpVqlTxv/d4PBo/frySkpL8x6ZMmRK86MIYIzcAANij0sXN9ddfr23btpU71q5dO+3atcv/nrklp3m8XkkUNwAAWK3Sxc3q1atDGEbkcZfwWAoAADtcFJl32rRpSk9PV1xcnNq2basNGzZU6roFCxbI4XCoe/fuoQ3wArD1AgAA9rA98y5cuFA5OTnKzc3Vpk2b1KJFC3Xu3FmHDh0673V79uzR0KFDdd1111kUaWDObAUHAADWsT3zTpkyRf369VPfvn3VtGlTTZ8+XVWqVNHs2bPPeY3X61WvXr00duxYNWjQwMJoK+90t1SUzZEAAHBpsbW48Xg82rhxY7l9qZxOp7KysrR+/fpzXvfEE0+oVq1auu+++370O9xut4qLi8u9rEC3FAAA9rA18xYWFsrr9SolJaXc8ZSUFOXn51d4zdq1a/Xyyy9r5syZlfqOiRMnKikpyf9KS0v7yXFXBsUNAAD2uKDMu2bNGv32t79VZmam9u/fL0maN2+e1q5dG9TgfujIkSPq3bu3Zs6cqeTk5EpdM3z4cBUVFflf+/btC2mMZdylp1rB6ZYCAMBaAS3iJ0l//etf1bt3b/Xq1UubN2+W2+2WJBUVFWnChAlatmxZpT8rOTlZUVFRKigoKHe8oKBAqampZ52/c+dO7dmzR127dvUf8/lOjZBER0dr27ZtatiwYblrXC6XXC5XpWMKFkZuAACwR8CZd9y4cZo+fbpmzpypmJgY//H27dtr06ZNAX1WbGysWrVqpVWrVvmP+Xw+rVq1SpmZmWedf+WVV+rzzz9XXl6e/3XbbbepY8eOysvLs+yRU2V42H4BAABbBDxys23bNl1//fVnHU9KStJ3330XcAA5OTnKzs5W69at1aZNG02dOlXHjh1T3759JUl9+vRR3bp1NXHiRMXFxalZs2blrq9evboknXXcbnRLAQBgj4CLm9TUVO3YsUPp6enljq9du/aC2rJ79Oihb775RqNHj1Z+fr4yMjK0fPly/yTjvXv3yukMv9EPD4v4AQBgi4CLm379+mnw4MGaPXu2HA6HDhw4oPXr12vo0KEaNWrUBQUxcOBADRw4sMK//di2D3PmzLmg7ww1FvEDAMAeARc3w4YNk8/n0w033KDjx4/r+uuvl8vl0tChQzVo0KBQxBiW6JYCAMAeARc3DodDjz/+uB555BHt2LFDR48eVdOmTVWtWrVQxBe23HRLAQBgi4CLmzKxsbFq2rRpMGOJKLSCAwBgj4CLm44dO8rhcJzz7++///5PCihS0C0FAIA9Ai5uMjIyyr0vKSlRXl6evvjiC2VnZwcrrrDHyA0AAPYIuLh59tlnKzw+ZswYHT169CcHFCkobgAAsEfQMu9vf/tbzZ49O1gfF/bolgIAwB5By7zr169XXFxcsD4u7LGIHwAA9gj4sdTtt99e7r0xRgcPHtQnn3xywYv4RSL/3lIs4gcAgKUCLm6SkpLKvXc6nWrcuLGeeOIJ3XTTTUELLNz5u6ViKG4AALBSQMWN1+tV37591bx5c9WoUSNUMUWE09sv0AoOAICVAhpWiIqK0k033XRBu39fauiWAgDAHgFn3mbNmmnXrl2hiCWinF7Ej+IGAAArBZx5x40bp6FDh+qdd97RwYMHVVxcXO6FUxi5AQDAHpWec/PEE0/oD3/4g2655RZJ0m233VZuGwZjjBwOh7xeb/CjDDPGGFrBAQCwSaWLm7Fjx+rBBx/UBx98EMp4IkLZIymJx1IAAFit0sWNMUaS1KFDh5AFEynKRm0kRm4AALBaQJn3fLuB4zTPGSM3sSziBwCApQJa56ZRo0Y/WuAcPnz4JwUUCdxnTCamIAQAwFoBFTdjx449a4VinI2tFwAAsE9Axc1dd92lWrVqhSqWiEEbOAAA9ql09uXxSuW5S0+1w9MpBQCA9Sqdfcu6pfDjGLkBAMA+lX4s5fP5fvwkSKK4AQDATmTfEDi9rxQ7ggMAYDWKmxBwM3IDAIBtyL4h4N9XilZwAAAsR/YNAXfJ991SMdxeAACsRvYNAUZuAACwD9k3BOiWAgDAPmTfEKBbCgAA+1DchAAjNwAA2IfsGwL+jTMpbgAAsBzZNwTYWwoAAPuQfUOAx1IAANiH7BsCtIIDAGAfsm8IuEu+n3PDIn4AAFiO7BsCbkZuAACwDdk3BE7PuWGdGwAArEZxEwJuWsEBALAN2TcEPN+3gtMtBQCA9ci+IUArOAAA9iH7hgCPpQAAsA/ZNwQYuQEAwD5k3xAoW8SPkRsAAKxH9g2BskX8YqNoBQcAwGoUNyHgH7lhhWIAACxH9g0B/5wbVigGAMByZN8QcLPODQAAtiH7BpnPZ1TiNZKYUAwAgB3IvkFWNt9GYuQGAAA7kH2DrGwBP4niBgAAO5B9g8xzZnHDhGIAACxH9g2yssdSsdFOORwOm6MBAODSQ3ETZO6SU51SLkZtAACwBRk4yFjADwAAe10UGXjatGlKT09XXFyc2rZtqw0bNpzz3JkzZ+q6665TjRo1VKNGDWVlZZ33fKuxgB8AAPayPQMvXLhQOTk5ys3N1aZNm9SiRQt17txZhw4dqvD81atX6+6779YHH3yg9evXKy0tTTfddJP2799vceQVc7MjOAAAtrI9A0+ZMkX9+vVT37591bRpU02fPl1VqlTR7NmzKzx//vz56t+/vzIyMnTllVdq1qxZ8vl8WrVqlcWRV6xs5MYVzaaZAADYwdbixuPxaOPGjcrKyvIfczqdysrK0vr16yv1GcePH1dJSYlq1qxZ4d/dbreKi4vLvULJw8gNAAC2sjUDFxYWyuv1KiUlpdzxlJQU5efnV+ozHnvsMdWpU6dcgXSmiRMnKikpyf9KS0v7yXGfD/tKAQBgr7DOwJMmTdKCBQv05ptvKi4ursJzhg8frqKiIv9r3759IY3J7X8sFda3FgCAsBVt55cnJycrKipKBQUF5Y4XFBQoNTX1vNc+88wzmjRpklauXKmrrrrqnOe5XC65XK6gxFsZPJYCAMBetmbg2NhYtWrVqtxk4LLJwZmZmee87qmnntKTTz6p5cuXq3Xr1laEWmluWsEBALCVrSM3kpSTk6Ps7Gy1bt1abdq00dSpU3Xs2DH17dtXktSnTx/VrVtXEydOlCT98Y9/1OjRo/Xaa68pPT3dPzenWrVqqlatmm2/o4y/WyqGbikAAOxge3HTo0cPffPNNxo9erTy8/OVkZGh5cuX+ycZ7927V07n6VGQl156SR6PR7/5zW/KfU5ubq7GjBljZegV8u8txcgNAAC2sL24kaSBAwdq4MCBFf5t9erV5d7v2bMn9AH9BO4S5twAAGAnMnCQebzfb5xJcQMAgC3IwEHmoRUcAABbkYGDjL2lAACwFxk4yBi5AQDAXmTgIGMRPwAA7EUGDjIW8QMAwF5k4CBzs4gfAAC2orgJMhbxAwDAXmTgIPOUnlrnhjk3AADYgwwcZG66pQAAsBUZOMjolgIAwF5k4CCjuAEAwF5k4CA7/ViKbikAAOxAcRNkrFAMAIC9yMBB5m8Fp7gBAMAWZOAgc5ecagVn5AYAAHuQgYOMkRsAAOxFBg4in8+oxGsksUIxAAB2IQMHUdmojcTeUgAA2IXiJojK2sAlRm4AALALGTiIPGcUNzFRDhsjAQDg0kVxE0Tu0tOdUg4HxQ0AAHaguAkitl4AAMB+ZOEgKptQzBo3AADYhywcRO4S9pUCAMBuFDdBxAJ+AADYjywcRP45N7SBAwBgG7JwEPm7pWK4rQAA2IUsHESM3AAAYD+ycBC5aQUHAMB2ZOEgKituaAUHAMA+ZOEgYhE/AADsRxYOotPFDevcAABgF4qbIOKxFAAA9iMLBxGPpQAAsB9ZOIg83lPr3NAKDgCAfcjCQeTfW4pF/AAAsA1ZOIj8u4IzcgMAgG3IwkHEnBsAAOxHFg6i091StIIDAGAXipsgYuQGAAD7kYWDiL2lAACwH1k4iNylp1rBWcQPAAD7kIWDiMdSAADYjywcRGWt4CziBwCAfcjCQXR6ET+6pQAAsAvFTRAxcgMAgP3IwkHEnBsAAOxHFg4iuqUAALAfWTiIPP4VirmtAADYhSwcRDyWAgDAfmThIGJvKQAA7EdxEyQ+n1Gpz0hi5AYAADuRhYOkrA1corgBAMBOZOEgKVvAT2JCMQAAdiILB4nbe6oN3OGQop0Om6MBAODSdVEUN9OmTVN6erri4uLUtm1bbdiw4bznL1q0SFdeeaXi4uLUvHlzLVu2zKJIz83fKRXllMNBcQMAgF1sL24WLlyonJwc5ebmatOmTWrRooU6d+6sQ4cOVXj+Rx99pLvvvlv33XefNm/erO7du6t79+764osvLI68PDdr3AAAcFFwGGOMnQG0bdtW11xzjV544QVJks/nU1pamgYNGqRhw4addX6PHj107NgxvfPOO/5j//3f/62MjAxNnz79R7+vuLhYSUlJKioqUmJiYtB+x9aDxbr5uTVKrubSJyOzgva5AAAgsPxt6zCDx+PRxo0blZV1uhhwOp3KysrS+vXrK7xm/fr15c6XpM6dO5/zfLfbreLi4nKvUGB1YgAALg62ZuLCwkJ5vV6lpKSUO56SkqL8/PwKr8nPzw/o/IkTJyopKcn/SktLC07wP1DqM6oSG6UqsSzgBwCAnSJ+mGH48OEqKiryv/bt2xeS72n18xra8kQXvZfTISSfDwAAKifazi9PTk5WVFSUCgoKyh0vKChQampqhdekpqYGdL7L5ZLL5QpOwAAA4KJn68hNbGysWrVqpVWrVvmP+Xw+rVq1SpmZmRVek5mZWe58SXrvvffOeT4AALi02DpyI0k5OTnKzs5W69at1aZNG02dOlXHjh1T3759JUl9+vRR3bp1NXHiREnS4MGD1aFDB02ePFm33nqrFixYoE8++UQzZsyw82cAAICLhO3FTY8ePfTNN99o9OjRys/PV0ZGhpYvX+6fNLx37145nacHmNq1a6fXXntNI0eO1IgRI3TFFVforbfeUrNmzez6CQAA4CJi+zo3VgvVOjcAACB0wmadGwAAgGCjuAEAABGF4gYAAEQUihsAABBRKG4AAEBEobgBAAARheIGAABEFIobAAAQUShuAABARLF9+wWrlS3IXFxcbHMkAACgssrydmU2VrjkipsjR45IktLS0myOBAAABOrIkSNKSko67zmX3N5SPp9PBw4cUEJCghwOR1A/u7i4WGlpadq3bx/7VoUQ99ka3GdrcJ+tw722RqjuszFGR44cUZ06dcptqF2RS27kxul0ql69eiH9jsTERP6LYwHuszW4z9bgPluHe22NUNznHxuxKcOEYgAAEFEobgAAQEShuAkil8ul3NxcuVwuu0OJaNxna3CfrcF9tg732hoXw32+5CYUAwCAyMbIDQAAiCgUNwAAIKJQ3AAAgIhCcQMAACIKxU2Apk2bpvT0dMXFxalt27basGHDec9ftGiRrrzySsXFxal58+ZatmyZRZGGt0Du88yZM3XdddepRo0aqlGjhrKysn70PxecEui/z2UWLFggh8Oh7t27hzbACBHoff7uu+80YMAA1a5dWy6XS40aNeJ/Oyoh0Ps8depUNW7cWPHx8UpLS9OQIUN08uRJi6INT3//+9/VtWtX1alTRw6HQ2+99daPXrN69Wq1bNlSLpdLl19+uebMmRPyOGVQaQsWLDCxsbFm9uzZ5ssvvzT9+vUz1atXNwUFBRWev27dOhMVFWWeeuops2XLFjNy5EgTExNjPv/8c4sjDy+B3ueePXuaadOmmc2bN5utW7eae+65xyQlJZmvv/7a4sjDS6D3uczu3btN3bp1zXXXXWe6detmTbBhLND77Ha7TevWrc0tt9xi1q5da3bv3m1Wr15t8vLyLI48vAR6n+fPn29cLpeZP3++2b17t1mxYoWpXbu2GTJkiMWRh5dly5aZxx9/3CxevNhIMm+++eZ5z9+1a5epUqWKycnJMVu2bDHPP/+8iYqKMsuXLw9pnBQ3AWjTpo0ZMGCA/73X6zV16tQxEydOrPD8O++809x6663ljrVt29b87ne/C2mc4S7Q+/xDpaWlJiEhwbz66quhCjEiXMh9Li0tNe3atTOzZs0y2dnZFDeVEOh9fumll0yDBg2Mx+OxKsSIEOh9HjBggOnUqVO5Yzk5OaZ9+/YhjTOSVKa4efTRR80vfvGLcsd69OhhOnfuHMLIjOGxVCV5PB5t3LhRWVlZ/mNOp1NZWVlav359hdesX7++3PmS1Llz53Oejwu7zz90/PhxlZSUqGbNmqEKM+xd6H1+4oknVKtWLd13331WhBn2LuQ+L1myRJmZmRowYIBSUlLUrFkzTZgwQV6v16qww86F3Od27dpp48aN/kdXu3bt0rJly3TLLbdYEvOlwq48eMltnHmhCgsL5fV6lZKSUu54SkqKvvrqqwqvyc/Pr/D8/Pz8kMUZ7i7kPv/QY489pjp16pz1XyicdiH3ee3atXr55ZeVl5dnQYSR4ULu865du/T++++rV69eWrZsmXbs2KH+/furpKREubm5VoQddi7kPvfs2VOFhYW69tprZYxRaWmpHnzwQY0YMcKKkC8Z58qDxcXFOnHihOLj40PyvYzcIKJMmjRJCxYs0Jtvvqm4uDi7w4kYR44cUe/evTVz5kwlJyfbHU5E8/l8qlWrlmbMmKFWrVqpR48eevzxxzV9+nS7Q4soq1ev1oQJE/Tiiy9q06ZNWrx4sZYuXaonn3zS7tAQBIzcVFJycrKioqJUUFBQ7nhBQYFSU1MrvCY1NTWg83Fh97nMM888o0mTJmnlypW66qqrQhlm2Av0Pu/cuVN79uxR165d/cd8Pp8kKTo6Wtu2bVPDhg1DG3QYupB/n2vXrq2YmBhFRUX5jzVp0kT5+fnyeDyKjY0Naczh6ELu86hRo9S7d2/df//9kqTmzZvr2LFjeuCBB/T444/L6eT/9w+Gc+XBxMTEkI3aSIzcVFpsbKxatWqlVatW+Y/5fD6tWrVKmZmZFV6TmZlZ7nxJeu+99855Pi7sPkvSU089pSeffFLLly9X69atrQg1rAV6n6+88kp9/vnnysvL879uu+02dezYUXl5eUpLS7My/LBxIf8+t2/fXjt27PAXj5K0fft21a5dm8LmHC7kPh8/fvysAqasoDRsuRg0tuXBkE5XjjALFiwwLpfLzJkzx2zZssU88MADpnr16iY/P98YY0zv3r3NsGHD/OevW7fOREdHm2eeecZs3brV5Obm0gpeCYHe50mTJpnY2FjzxhtvmIMHD/pfR44csesnhIVA7/MP0S1VOYHe571795qEhAQzcOBAs23bNvPOO++YWrVqmXHjxtn1E8JCoPc5NzfXJCQkmL/85S9m165d5v/+7/9Mw4YNzZ133mnXTwgLR44cMZs3bzabN282ksyUKVPM5s2bzb///W9jjDHDhg0zvXv39p9f1gr+yCOPmK1bt5pp06bRCn4xev75581ll11mYmNjTZs2bcw//vEP/986dOhgsrOzy53/+uuvm0aNGpnY2Fjzi1/8wixdutTiiMNTIPf55z//uZF01is3N9f6wMNMoP8+n4nipvICvc8fffSRadu2rXG5XKZBgwZm/PjxprS01OKow08g97mkpMSMGTPGNGzY0MTFxZm0tDTTv39/8+2331ofeBj54IMPKvzf27J7m52dbTp06HDWNRkZGSY2NtY0aNDAvPLKKyGP02EM428AACByMOcGAABEFIobAAAQUShuAABARKG4AQAAEYXiBgAARBSKGwAAEFEobgAAQEShuAEAABGF4gZAOXPmzFH16tXtDuOCORwOvfXWW+c955577lH37t0tiQeA9ShugAh0zz33yOFwnPXasWOH3aFpzpw5/nicTqfq1aunvn376tChQ0H5/IMHD+rmm2+WJO3Zs0cOh0N5eXnlznnuuec0Z86coHzfuYwZM8b/O6OiopSWlqYHHnhAhw8fDuhzKMSAwEXbHQCA0OjSpYteeeWVcsd+9rOf2RRNeYmJidq2bZt8Pp8+/fRT9e3bVwcOHNCKFSt+8menpqb+6DlJSUk/+Xsq4xe/+IVWrlwpr9errVu36t5771VRUZEWLlxoyfcDlypGboAI5XK5lJqaWu4VFRWlKVOmqHnz5qpatarS0tLUv39/HT169Jyf8+mnn6pjx45KSEhQYmKiWrVqpU8++cT/97Vr1+q6665TfHy80tLS9PDDD+vYsWPnjc3hcCg1NVV16tTRzTffrIcfflgrV67UiRMn5PP59MQTT6hevXpyuVzKyMjQ8uXL/dd6PB4NHDhQtWvXVlxcnH7+859r4sSJ5T677LFU/fr1JUlXX321HA6H/ud//kdS+dGQGTNmqE6dOvL5fOVi7Natm+69917/+7ffflstW7ZUXFycGjRooLFjx6q0tPS8vzM6OlqpqamqW7eusrKydMcdd+i9997z/93r9eq+++5T/fr1FR8fr8aNG+u5557z/33MmDF69dVX9fbbb/tHgVavXi1J2rdvn+68805Vr15dNWvWVLdu3bRnz57zxgNcKihugEuM0+nUn/70J3355Zd69dVX9f777+vRRx895/m9evVSvXr19PHHH2vjxo0aNmyYYmJiJEk7d+5Uly5d9Otf/1qfffaZFi5cqLVr12rgwIEBxRQfHy+fz6fS0lI999xzmjx5sp555hl99tln6ty5s2677Tb961//kiT96U9/0pIlS/T6669r27Ztmj9/vtLT0yv83A0bNkiSVq5cqYMHD2rx4sVnnXPHHXfoP//5jz744AP/scOHD2v58uXq1auXJGnNmjXq06ePBg8erC1btujPf/6z5syZo/Hjx1f6N+7Zs0crVqxQbGys/5jP51O9evW0aNEibdmyRaNHj9aIESP0+uuvS5KGDh2qO++8U126dNHBgwd18OBBtWvXTiUlJercubMSEhK0Zs0arVu3TtWqVVOXLl3k8XgqHRMQsUK+7zgAy2VnZ5uoqChTtWpV/+s3v/lNhecuWrTI/Nd//Zf//SuvvGKSkpL87xMSEsycOXMqvPa+++4zDzzwQLlja9asMU6n05w4caLCa374+du3bzeNGjUyrVu3NsYYU6dOHTN+/Phy11xzzTWmf//+xhhjBg0aZDp16mR8Pl+Fny/JvPnmm8YYY3bv3m0kmc2bN5c7Jzs723Tr1s3/vlu3bubee+/1v//zn/9s6tSpY7xerzHGmBtuuMFMmDCh3GfMmzfP1K5du8IYjDEmNzfXOJ1OU7VqVRMXF2ckGUlmypQp57zGGGMGDBhgfv3rX58z1rLvbty4cbl74Ha7TXx8vFmxYsV5Px+4FDDnBohQHTt21EsvveR/X7VqVUmnRjEmTpyor776SsXFxSotLdXJkyd1/PhxValS5azPycnJ0f3336958+b5H600bNhQ0qlHVp999pnmz5/vP98YI5/Pp927d6tJkyYVxlZUVKRq1arJ5/Pp5MmTuvbaazVr1iwVFxfrwIEDat++fbnz27dvr08//VTSqUdKN954oxo3bqwuXbrol7/8pW666aafdK969eqlfv366cUXX5TL5dL8+fN11113yel0+n/nunXryo3UeL3e8943SWrcuLGWLFmikydP6n//93+Vl5enQYMGlTtn2rRpmj17tvbu3asTJ07I4/EoIyPjvPF++umn2rFjhxISEsodP3nypHbu3HkBdwCILBQ3QISqWrWqLr/88nLH9uzZo1/+8pd66KGHNH78eNWsWVNr167VfffdJ4/HU2GSHjNmjHr27KmlS5fq3XffVW5urhYsWKBf/epXOnr0qH73u9/p4YcfPuu6yy677JyxJSQkaNOmTXI6napdu7bi4+MlScXFxT/6u1q2bKndu3fr3Xff1cqVK3XnnXcqKytLb7zxxo9eey5du3aVMUZLly7VNddcozVr1ujZZ5/1//3o0aMaO3asbr/99rOujYuLO+fnxsbG+v8zmDRpkm699VaNHTtWTz75pCRpwYIFGjp0qCZPnqzMzEwlJCTo6aef1j//+c/zxnv06FG1atWqXFFZ5mKZNA7YieIGuIRs3LhRPp9PkydP9o9KlM3vOJ9GjRqpUaNGGjJkiO6++2698sor+tWvfqWWLVtqy5YtZxVRP8bpdFZ4TWJiourUqaN169apQ4cO/uPr1q1TmzZtyp3Xo0cP9ejRQ7/5zW/UpUsXHT58WDVr1iz3eWXzW7xe73njiYuL0+2336758+drx44daty4sVq2bOn/e8uWLbVt27aAf+cPjRw5Up06ddJDDz3k/53t2rVT//79/ef8cOQlNjb2rPhbtmyphQsXqlatWkpMTPxJMQGRiAnFwCXk8ssvV0lJiZ5//nnt2rVL8+bN0/Tp0895/okTJzRw4ECtXr1a//73v7Vu3Tp9/PHH/sdNjz32mD766CMNHDhQeXl5+te//qW333474AnFZ3rkkUf0xz/+UQsXLtS2bds0bNgw5eXlafDgwZKkKVOm6C9/+Yu++uorbd++XYsWLVJqamqFCw/WqlVL8fHxWr58uQoKClRUVHTO7+3Vq5eWLl2q2bNn+ycSlxk9erTmzp2rsWPH6ssvv9TWrVu1YMECjRw5MqDflpmZqauuukoTJkyQJF1xxRX65JNPtGLFCm3fvl2jRo3Sxx9/XO6a9PR0ffbZZ9q2bZsKCwtVUlKiXr16KTk5Wd26ddOaNWu0e/durV69Wg8//LC+/vrrgGICIpLdk34ABF9Fk1DLTJkyxdSuXdvEx8ebzp07m7lz5xpJ5ttvvzXGlJ/w63a7zV133WXS0tJMbGysqVOnjhk4cGC5ycIbNmwwN954o6lWrZqpWrWqueqqq86aEHymH04o/iGv12vGjBlj6tata2JiYkyLFi3Mu+++6//7jBkzTEZGhqlatapJTEw0N9xwg9m0aZP/7zpjQrExxsycOdOkpaUZp9NpOnTocM774/V6Te3atY0ks3PnzrPiWr58uWnXrp2Jj483iYmJpk2bNmbGjBnn/B25ubmmRYsWZx3/y1/+Ylwul9m7d685efKkueeee0xSUpKpXr26eeihh8ywYcPKXXfo0CH//ZVkPvjgA2OMMQcPHjR9+vQxycnJxuVymQYNGph+/fqZoqKic8YEXCocxhhjb3kFAAAQPDyWAgAAEYXiBgAARBSKGwAAEFEobgAAQEShuAEAABGF4gYAAEQUihsAABBRKG4AAEBEobgBAAARheIGAABEFIobAAAQUf4feEB+wcsAtcgAAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "fpr,tpr, threshold = roc_curve(Y_test_dt, decision_tree_predictions)\n",
        "disp = RocCurveDisplay(fpr=fpr, tpr=tpr)\n",
        "disp.plot()\n",
        "plt.show()"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "collapsed_sections": [
        "n62_Iy9439Uk",
        "ElTPbEZ7HP-r",
        "upPOeNcZ3wG_",
        "RKbyrDU1PVQj"
      ],
      "provenance": []
    },
    "kaggle": {
      "accelerator": "none",
      "dataSources": [
        {
          "datasetId": 4030196,
          "sourceId": 7009925,
          "sourceType": "datasetVersion"
        }
      ],
      "dockerImageVersionId": 30646,
      "isGpuEnabled": false,
      "isInternetEnabled": true,
      "language": "python",
      "sourceType": "notebook"
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.10.13"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}