bank_marketing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load Libraries and Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total number of instances : 45211\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<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>age</th>\n",
       "      <th>job</th>\n",
       "      <th>marital</th>\n",
       "      <th>education</th>\n",
       "      <th>default</th>\n",
       "      <th>balance</th>\n",
       "      <th>housing</th>\n",
       "      <th>loan</th>\n",
       "      <th>contact</th>\n",
       "      <th>day</th>\n",
       "      <th>month</th>\n",
       "      <th>duration</th>\n",
       "      <th>campaign</th>\n",
       "      <th>pdays</th>\n",
       "      <th>previous</th>\n",
       "      <th>poutcome</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>58</td>\n",
       "      <td>management</td>\n",
       "      <td>married</td>\n",
       "      <td>tertiary</td>\n",
       "      <td>no</td>\n",
       "      <td>2143</td>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "      <td>unknown</td>\n",
       "      <td>5</td>\n",
       "      <td>may</td>\n",
       "      <td>261</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>unknown</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>44</td>\n",
       "      <td>technician</td>\n",
       "      <td>single</td>\n",
       "      <td>secondary</td>\n",
       "      <td>no</td>\n",
       "      <td>29</td>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "      <td>unknown</td>\n",
       "      <td>5</td>\n",
       "      <td>may</td>\n",
       "      <td>151</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>unknown</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>33</td>\n",
       "      <td>entrepreneur</td>\n",
       "      <td>married</td>\n",
       "      <td>secondary</td>\n",
       "      <td>no</td>\n",
       "      <td>2</td>\n",
       "      <td>yes</td>\n",
       "      <td>yes</td>\n",
       "      <td>unknown</td>\n",
       "      <td>5</td>\n",
       "      <td>may</td>\n",
       "      <td>76</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>unknown</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>47</td>\n",
       "      <td>blue-collar</td>\n",
       "      <td>married</td>\n",
       "      <td>unknown</td>\n",
       "      <td>no</td>\n",
       "      <td>1506</td>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "      <td>unknown</td>\n",
       "      <td>5</td>\n",
       "      <td>may</td>\n",
       "      <td>92</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>unknown</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>33</td>\n",
       "      <td>unknown</td>\n",
       "      <td>single</td>\n",
       "      <td>unknown</td>\n",
       "      <td>no</td>\n",
       "      <td>1</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>unknown</td>\n",
       "      <td>5</td>\n",
       "      <td>may</td>\n",
       "      <td>198</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>unknown</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age           job  marital  education default  balance housing loan  \\\n",
       "0   58    management  married   tertiary      no     2143     yes   no   \n",
       "1   44    technician   single  secondary      no       29     yes   no   \n",
       "2   33  entrepreneur  married  secondary      no        2     yes  yes   \n",
       "3   47   blue-collar  married    unknown      no     1506     yes   no   \n",
       "4   33       unknown   single    unknown      no        1      no   no   \n",
       "\n",
       "   contact  day month  duration  campaign  pdays  previous poutcome   y  \n",
       "0  unknown    5   may       261         1     -1         0  unknown  no  \n",
       "1  unknown    5   may       151         1     -1         0  unknown  no  \n",
       "2  unknown    5   may        76         1     -1         0  unknown  no  \n",
       "3  unknown    5   may        92         1     -1         0  unknown  no  \n",
       "4  unknown    5   may       198         1     -1         0  unknown  no  "
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "from collections import defaultdict\n",
    "\n",
    "df = pd.read_csv('bank-full.csv', delimiter=';')\n",
    "\n",
    "print(f\"Total number of instances : {len(df)}\")\n",
    "\n",
    "for i in df.columns :\n",
    "    num_nan = df[i].isnull().sum()\n",
    "    if num_nan > 0:\n",
    "        print(f\"{i} column has {num_nan} NaN values\")\n",
    "\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dataset Description"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Class Labels : ['yes', 'no']\n",
      "No. of Inst. : [5289, 39922]\n",
      "\n",
      "Total number of features : 16\n"
     ]
    }
   ],
   "source": [
    "y = list(df['y'])\n",
    "instances_per_class = {'yes' : y.count('yes'), 'no' : y.count('no')}\n",
    "\n",
    "plt.bar([0, 1], instances_per_class.values(), color='green')\n",
    "plt.xlabel('Class')\n",
    "plt.ylabel('No. of Instances')\n",
    "plt.title('Class Imbalance Check')\n",
    "plt.xticks([0, 1], ['Subscribed', 'Not Subscribed'])\n",
    "plt.show()\n",
    "\n",
    "print(f\"Class Labels : {list(instances_per_class.keys())}\\nNo. of Inst. : {list(instances_per_class.values())}\\n\\nTotal number of features : {len(df.columns)-1}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'age': 77,\n",
       " 'job': 12,\n",
       " 'marital': 3,\n",
       " 'education': 4,\n",
       " 'default': 2,\n",
       " 'balance': 7168,\n",
       " 'housing': 2,\n",
       " 'loan': 2,\n",
       " 'contact': 3,\n",
       " 'day': 31,\n",
       " 'month': 12,\n",
       " 'duration': 1573,\n",
       " 'campaign': 48,\n",
       " 'pdays': 559,\n",
       " 'previous': 41,\n",
       " 'poutcome': 4}"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = df['y']\n",
    "X = df.drop('y', axis=1)\n",
    "\n",
    "unique_val_count = {}\n",
    "\n",
    "for col in X.columns:\n",
    "    unique_values = np.unique(X[col])\n",
    "    unique_val_count[col] = len(unique_values)\n",
    "\n",
    "unique_val_count"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Categorical to Numerical"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of Categorical columns : 10\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<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>age</th>\n",
       "      <th>job</th>\n",
       "      <th>marital</th>\n",
       "      <th>education</th>\n",
       "      <th>default</th>\n",
       "      <th>balance</th>\n",
       "      <th>housing</th>\n",
       "      <th>loan</th>\n",
       "      <th>contact</th>\n",
       "      <th>day</th>\n",
       "      <th>month</th>\n",
       "      <th>duration</th>\n",
       "      <th>campaign</th>\n",
       "      <th>pdays</th>\n",
       "      <th>previous</th>\n",
       "      <th>poutcome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>58</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2143</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>261</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>44</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>151</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>33</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>76</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>47</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1506</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>92</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>33</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</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",
       "      <td>198</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>35</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>231</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>139</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>28</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>447</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>217</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>42</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>380</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>58</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>121</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>50</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>43</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>593</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>55</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age  job  marital  education  default  balance  housing  loan  contact  \\\n",
       "0   58    0        0          0        0     2143        0     0        0   \n",
       "1   44    1        1          1        0       29        0     0        0   \n",
       "2   33    2        0          1        0        2        0     1        0   \n",
       "3   47    3        0          2        0     1506        0     0        0   \n",
       "4   33    4        1          2        0        1        1     0        0   \n",
       "5   35    0        0          0        0      231        0     0        0   \n",
       "6   28    0        1          0        0      447        0     1        0   \n",
       "7   42    2        2          0        1        2        0     0        0   \n",
       "8   58    5        0          3        0      121        0     0        0   \n",
       "9   43    1        1          1        0      593        0     0        0   \n",
       "\n",
       "   day  month  duration  campaign  pdays  previous  poutcome  \n",
       "0    0      0       261         1     -1         0         0  \n",
       "1    0      0       151         1     -1         0         0  \n",
       "2    0      0        76         1     -1         0         0  \n",
       "3    0      0        92         1     -1         0         0  \n",
       "4    0      0       198         1     -1         0         0  \n",
       "5    0      0       139         1     -1         0         0  \n",
       "6    0      0       217         1     -1         0         0  \n",
       "7    0      0       380         1     -1         0         0  \n",
       "8    0      0        50         1     -1         0         0  \n",
       "9    0      0        55         1     -1         0         0  "
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "unique_val_count = {}\n",
    "\n",
    "cat = list(X.columns)\n",
    "cat.remove('age')\n",
    "cat.remove('balance')\n",
    "cat.remove('duration')\n",
    "cat.remove('pdays')\n",
    "cat.remove('campaign')\n",
    "cat.remove('previous')\n",
    "\n",
    "for col in X.columns:\n",
    "    if col not in cat:\n",
    "        continue\n",
    "    unique_values = np.unique(X[col])\n",
    "    temp = defaultdict(int)\n",
    "    for val in X[col]:\n",
    "        temp[val] += 1\n",
    "    unique_val_count[col] = temp\n",
    "\n",
    "print(f\"Number of Categorical columns : {len(unique_val_count.keys())}\")\n",
    "\n",
    "for col, attr in unique_val_count.items():\n",
    "    map_val = {}\n",
    "    count = 0\n",
    "    for key in attr.keys():\n",
    "        map_val[key] = count\n",
    "        count += 1\n",
    "    X[col] = X[col].map(map_val)\n",
    "\n",
    "X.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'day', 'month', 'poutcome'])"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "unique_val_count.keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Visualization of Categorical Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "cross_tables = []\n",
    "for i in unique_val_count.keys():\n",
    "    cross_tables.append(pd.crosstab(X[i], y))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Job type vs Subscription Status"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "width = 0.25\n",
    "plt.bar(np.arange(0, len(cross_tables[0])), cross_tables[0]['no'], color='blue', width=width, edgecolor='black', label='Not Subscribed')\n",
    "plt.bar(np.arange(0, len(cross_tables[0]))+width, cross_tables[0]['yes'], width=width, color='green', edgecolor='black', label='Subscribed')\n",
    "plt.xticks(np.arange(0, len(cross_tables[0])) + width/2, range(len(cross_tables[0])))\n",
    "plt.xlabel('Job')\n",
    "plt.ylabel('Number of Instances')\n",
    "plt.ylim([0, max((list(cross_tables[0]['no'])) + list(cross_tables[0]['yes']))+100])\n",
    "plt.legend()\n",
    "plt.title('Job type vs Subscription Status')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Marital Status vs Subscription Status"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "width = 0.25\n",
    "plt.bar(np.arange(0, len(cross_tables[1])), cross_tables[1]['no'], color='blue', width=width, edgecolor='black', label='Not Subscribed')\n",
    "plt.bar(np.arange(0, len(cross_tables[1]))+width, cross_tables[1]['yes'], width=width, color='green', edgecolor='black', label='Subscribed')\n",
    "plt.xticks(np.arange(0, len(cross_tables[1])) + width/2, unique_val_count['marital'].keys())\n",
    "plt.xlabel('Marital Status')\n",
    "plt.ylabel('Number of Instances')\n",
    "plt.ylim([0, max((list(cross_tables[1]['no'])) + list(cross_tables[1]['yes']))+1000])\n",
    "plt.legend()\n",
    "plt.title('Marital Status vs Subscription Status')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Education vs Subscription Status"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "width = 0.25\n",
    "plt.bar(np.arange(0, len(cross_tables[2])), cross_tables[2]['no'], color='blue', width=width, edgecolor='black', label='Not Subscribed')\n",
    "plt.bar(np.arange(0, len(cross_tables[2]))+width, cross_tables[2]['yes'], width=width, color='green', edgecolor='black', label='Subscribed')\n",
    "plt.xticks(np.arange(0, len(cross_tables[2])) + width/2, unique_val_count['education'].keys())\n",
    "plt.xlabel('Education')\n",
    "plt.ylabel('Number of Instances')\n",
    "plt.ylim([0, max((list(cross_tables[2]['no'])) + list(cross_tables[2]['yes']))+1000])\n",
    "plt.legend()\n",
    "plt.title('Education Level vs Subscription Status')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Defaulter vs Subscription Status"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "width = 0.25\n",
    "plt.bar(np.arange(0, len(cross_tables[3])), cross_tables[3]['no'], color='blue', width=width, edgecolor='black', label='Not Subscribed')\n",
    "plt.bar(np.arange(0, len(cross_tables[3]))+width, cross_tables[3]['yes'], width=width, color='green', edgecolor='black', label='Subscribed')\n",
    "plt.xticks(np.arange(0, len(cross_tables[3])) + width/2, unique_val_count['default'].keys())\n",
    "plt.xlabel('Defaulter')\n",
    "plt.ylabel('Number of Instances')\n",
    "plt.ylim([0, max((list(cross_tables[3]['no'])) + list(cross_tables[3]['yes']))+1000])\n",
    "plt.legend()\n",
    "plt.title('Defaulter Status vs Subscription Status')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Housing Status vs Subscription Status"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "width = 0.25\n",
    "plt.bar(np.arange(0, len(cross_tables[4])), cross_tables[4]['no'], color='blue', width=width, edgecolor='black', label='Not Subscribed')\n",
    "plt.bar(np.arange(0, len(cross_tables[4]))+width, cross_tables[4]['yes'], width=width, color='green', edgecolor='black', label='Subscribed')\n",
    "plt.xticks(np.arange(0, len(cross_tables[4])) + width/2, unique_val_count['housing'].keys())\n",
    "plt.xlabel('Housing Status')\n",
    "plt.ylabel('Number of Instances')\n",
    "plt.ylim([0, max((list(cross_tables[4]['no'])) + list(cross_tables[4]['yes']))+1000])\n",
    "plt.legend()\n",
    "plt.title('Housing Status vs Subscription Status')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Loan status vs Subscription Status"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "width = 0.25\n",
    "plt.bar(np.arange(0, len(cross_tables[5])), cross_tables[5]['no'], color='blue', width=width, edgecolor='black', label='Not Subscribed')\n",
    "plt.bar(np.arange(0, len(cross_tables[5]))+width, cross_tables[5]['yes'], width=width, color='green', edgecolor='black', label='Subscribed')\n",
    "plt.xticks(np.arange(0, len(cross_tables[5])) + width/2, unique_val_count['loan'].keys())\n",
    "plt.xlabel('Loan Status')\n",
    "plt.ylabel('Number of Instances')\n",
    "plt.ylim([0, max((list(cross_tables[5]['no'])) + list(cross_tables[5]['yes']))+1000])\n",
    "plt.legend()\n",
    "plt.title('Loan Status vs Subscription Status')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Contact type vs Subscription Status"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "width = 0.25\n",
    "plt.bar(np.arange(0, len(cross_tables[6])), cross_tables[6]['no'], color='blue', width=width, edgecolor='black', label='Not Subscribed')\n",
    "plt.bar(np.arange(0, len(cross_tables[6]))+width, cross_tables[6]['yes'], width=width, color='green', edgecolor='black', label='Subscribed')\n",
    "plt.xticks(np.arange(0, len(cross_tables[6])) + width/2, unique_val_count['contact'].keys())\n",
    "plt.xlabel('Contact Type')\n",
    "plt.ylabel('Number of Instances')\n",
    "plt.ylim([0, max((list(cross_tables[6]['no'])) + list(cross_tables[6]['yes']))+1000])\n",
    "plt.legend()\n",
    "plt.title('Contact type vs Subscription Status')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Last Day contacted vs Subscription Status"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "width = 0.25\n",
    "plt.bar(np.arange(0, len(cross_tables[7])), cross_tables[7]['no'], color='blue', width=width, edgecolor='black', label='Not Subscribed')\n",
    "plt.bar(np.arange(0, len(cross_tables[7]))+width, cross_tables[7]['yes'], width=width, color='green', edgecolor='black', label='Subscribed')\n",
    "plt.xticks(np.arange(0, len(cross_tables[7])) + width/2, unique_val_count['day'].keys())\n",
    "plt.xlabel('Day of the Month')\n",
    "plt.ylabel('Number of Instances')\n",
    "plt.ylim([0, max((list(cross_tables[7]['no'])) + list(cross_tables[7]['yes']))+1000])\n",
    "plt.legend()\n",
    "plt.title('Last Contact Day vs Subscription Status')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Last contact month vs Subscription Status"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "width = 0.25\n",
    "plt.bar(np.arange(0, len(cross_tables[8])), cross_tables[8]['no'], color='blue', width=width, edgecolor='black', label='Not Subscribed')\n",
    "plt.bar(np.arange(0, len(cross_tables[8]))+width, cross_tables[8]['yes'], width=width, color='green', edgecolor='black', label='Subscribed')\n",
    "plt.xticks(np.arange(0, len(cross_tables[8])) + width/2, unique_val_count['month'].keys())\n",
    "plt.xlabel('Month name')\n",
    "plt.ylabel('Number of Instances')\n",
    "plt.ylim([0, max((list(cross_tables[8]['no'])) + list(cross_tables[8]['yes']))+1000])\n",
    "plt.legend()\n",
    "plt.title('Last Contact Month vs Subscription Status')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Outcome of previous campaign vs Subscription Status"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "width = 0.25\n",
    "plt.bar(np.arange(0, len(cross_tables[9])), cross_tables[9]['no'], color='blue', width=width, edgecolor='black', label='Not Subscribed')\n",
    "plt.bar(np.arange(0, len(cross_tables[9]))+width, cross_tables[9]['yes'], width=width, color='green', edgecolor='black', label='Subscribed')\n",
    "plt.xticks(np.arange(0, len(cross_tables[9])) + width/2, unique_val_count['poutcome'].keys())\n",
    "plt.xlabel('Outcome')\n",
    "plt.ylabel('Number of Instances')\n",
    "plt.ylim([0, max((list(cross_tables[9]['no'])) + list(cross_tables[9]['yes']))+1000])\n",
    "plt.legend()\n",
    "plt.title('Outcome of Prev Campaign vs Subscription Status')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Visualization after PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.decomposition import PCA\n",
    "\n",
    "y = y.map({'yes' : 1, 'no' : 0})\n",
    "X = np.array(X)\n",
    "y = np.array(y)\n",
    "\n",
    "X_pca = PCA(n_components=2).fit_transform(X)\n",
    "\n",
    "x_no = []\n",
    "x_yes = []\n",
    "for i in range(len(y)):\n",
    "    if y[i] == 0:\n",
    "        x_no.append(X_pca[i])\n",
    "    else:\n",
    "        x_yes.append(X_pca[i])\n",
    "\n",
    "x_no = np.array(x_no)\n",
    "x_yes = np.array(x_yes)\n",
    "\n",
    "plt.scatter(x_no[:,0], x_no[:,1], color='blue', label='Not Subscribed')\n",
    "plt.scatter(x_yes[:,0], x_yes[:,1], color='green', label='Subscribed')\n",
    "plt.xlabel('PC-1')\n",
    "plt.ylabel('PC-2')\n",
    "plt.legend()\n",
    "plt.title('PCA for 2D visualization')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_pca = PCA(n_components=3).fit_transform(X)\n",
    "\n",
    "x_no = []\n",
    "x_yes = []\n",
    "for i in range(len(y)):\n",
    "    if y[i] == 0:\n",
    "        x_no.append(X_pca[i])\n",
    "    else:\n",
    "        x_yes.append(X_pca[i])\n",
    "\n",
    "x_no = np.array(x_no)\n",
    "x_yes = np.array(x_yes)\n",
    "\n",
    "plt.figure()\n",
    "ax = plt.axes(projection='3d')\n",
    "ax.scatter3D(x_no[:,0], x_no[:,1], x_no[:,2], cmap='Blues')\n",
    "ax.scatter3D(x_yes[:,0], x_yes[:,1], x_yes[:,2], cmap='Reds')\n",
    "ax.set_xlabel('PC-1')\n",
    "ax.set_ylabel('PC-2')\n",
    "ax.set_zlabel('PC-3')\n",
    "plt.legend(['Not Subscribed', 'Subscribed'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Normalization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 1.60696496, -1.13808302, -0.74066636, -1.08356042, -0.13548989,\n",
       "         0.25641925, -0.89391541, -0.43680347, -1.41210845, -1.55754973,\n",
       "        -0.92484102,  0.0110161 , -0.56935064, -0.41145311, -0.25194037,\n",
       "        -0.41528717],\n",
       "       [ 0.28852927, -0.81532633,  0.70245839, -0.05154268, -0.13548989,\n",
       "        -0.43789469, -0.89391541, -0.43680347, -1.41210845, -1.55754973,\n",
       "        -0.92484102, -0.41612696, -0.56935064, -0.41145311, -0.25194037,\n",
       "        -0.41528717]])"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "X_scaled = StandardScaler().fit_transform(X)\n",
    "X_scaled[:2]\n",
    "#scaler = MinMaxScaler().fit(train)\n",
    "#X_scaled = scaler.transform(train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Splitting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train size : 36168\tValidation size : 9043\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_val, y_train, y_val = train_test_split(X_scaled, y, test_size=0.2)\n",
    "print(f\"Train size : {len(X_train)}\\tValidation size : {len(X_val)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model Selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score, f1_score, roc_auc_score, roc_curve, RocCurveDisplay\n",
    "import copy, pickle\n",
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "models_path = 'D:\\IIT JODHPUR\\Semester-3\\ML for Business\\Course Project\\\\best_models'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "C : 0.001\tAccuracy : 0.826\tF1 Score : 0.850\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "C : 0.010\tAccuracy : 0.826\tF1 Score : 0.850\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "C : 0.100\tAccuracy : 0.826\tF1 Score : 0.850\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "C : 1.000\tAccuracy : 0.826\tF1 Score : 0.850\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "C : 10.000\tAccuracy : 0.826\tF1 Score : 0.850\n",
      "Best Model based on Validation Accuracy : {'C': 0.001, 'acc': 0.8264956319805374, 'f1': 0.8503038288499886, 'model': LogisticRegression(C=0.001, class_weight='balanced')}\n",
      "Validation Accuracy : 0.8264956319805374\n",
      "Test F1-Score : 0.8503038288499886\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<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>C</th>\n",
       "      <th>cv_score_fold1</th>\n",
       "      <th>cv_score_fold2</th>\n",
       "      <th>cv_score_fold3</th>\n",
       "      <th>cv_score_fold4</th>\n",
       "      <th>cv_score_fold5</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>accuracy</th>\n",
       "      <th>f1_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.001</td>\n",
       "      <td>0.856906</td>\n",
       "      <td>0.826698</td>\n",
       "      <td>0.840080</td>\n",
       "      <td>0.284119</td>\n",
       "      <td>0.669985</td>\n",
       "      <td>0.695557</td>\n",
       "      <td>0.216358</td>\n",
       "      <td>0.826496</td>\n",
       "      <td>0.850304</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.010</td>\n",
       "      <td>0.863320</td>\n",
       "      <td>0.811436</td>\n",
       "      <td>0.838310</td>\n",
       "      <td>0.276709</td>\n",
       "      <td>0.687680</td>\n",
       "      <td>0.695491</td>\n",
       "      <td>0.217920</td>\n",
       "      <td>0.825832</td>\n",
       "      <td>0.849926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.100</td>\n",
       "      <td>0.865310</td>\n",
       "      <td>0.806348</td>\n",
       "      <td>0.841739</td>\n",
       "      <td>0.254369</td>\n",
       "      <td>0.686021</td>\n",
       "      <td>0.690757</td>\n",
       "      <td>0.226752</td>\n",
       "      <td>0.826053</td>\n",
       "      <td>0.850095</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.870176</td>\n",
       "      <td>0.809998</td>\n",
       "      <td>0.864189</td>\n",
       "      <td>0.278257</td>\n",
       "      <td>0.671422</td>\n",
       "      <td>0.698808</td>\n",
       "      <td>0.222123</td>\n",
       "      <td>0.826164</td>\n",
       "      <td>0.850179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10.000</td>\n",
       "      <td>0.869402</td>\n",
       "      <td>0.806791</td>\n",
       "      <td>0.865295</td>\n",
       "      <td>0.276266</td>\n",
       "      <td>0.673413</td>\n",
       "      <td>0.698233</td>\n",
       "      <td>0.222560</td>\n",
       "      <td>0.826164</td>\n",
       "      <td>0.850179</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        C  cv_score_fold1  cv_score_fold2  cv_score_fold3  cv_score_fold4  \\\n",
       "0   0.001        0.856906        0.826698        0.840080        0.284119   \n",
       "1   0.010        0.863320        0.811436        0.838310        0.276709   \n",
       "2   0.100        0.865310        0.806348        0.841739        0.254369   \n",
       "3   1.000        0.870176        0.809998        0.864189        0.278257   \n",
       "4  10.000        0.869402        0.806791        0.865295        0.276266   \n",
       "\n",
       "   cv_score_fold5      mean       std  accuracy  f1_score  \n",
       "0        0.669985  0.695557  0.216358  0.826496  0.850304  \n",
       "1        0.687680  0.695491  0.217920  0.825832  0.849926  \n",
       "2        0.686021  0.690757  0.226752  0.826053  0.850095  \n",
       "3        0.671422  0.698808  0.222123  0.826164  0.850179  \n",
       "4        0.673413  0.698233  0.222560  0.826164  0.850179  "
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "C = [0.001, 0.01, 0.1, 1, 10]\n",
    "ALL_MODELS = []\n",
    "\n",
    "best_model = {'C' : 0, 'acc' : 0, 'f1' : 0, 'model' : None}\n",
    "for c in C:\n",
    "    model = LogisticRegression(C=c, class_weight='balanced')\n",
    "    scores = cross_val_score(model, X, y, cv=5)\n",
    "    model.fit(X_train, y_train)\n",
    "    y_pred = model.predict(X_val)\n",
    "    acc = accuracy_score(y_true=y_val, y_pred=y_pred)\n",
    "    f1 = f1_score(y_true=y_val, y_pred=y_pred, average='weighted')\n",
    "    print(f\"C : {c:.3f}\\tAccuracy : {acc:.3f}\\tF1 Score : {f1:.3f}\")\n",
    "    ALL_MODELS.append(pd.Series({'C' : c, 'cv_score_fold1' : scores[0], 'cv_score_fold2' : scores[1], 'cv_score_fold3' : scores[2], 'cv_score_fold4' : scores[3], 'cv_score_fold5' : scores[4], 'mean' : scores.mean(), 'std' : scores.std(), 'accuracy' : acc, 'f1_score' : f1}))\n",
    "    if best_model['f1'] < f1:\n",
    "        best_model = {'C' : c, 'acc' : acc, 'f1' : f1, 'model' : copy.deepcopy(model)}\n",
    "\n",
    "y_pred = best_model['model'].predict(X_val)\n",
    "print(f\"Best Model based on Validation Accuracy : {best_model}\\nValidation Accuracy : {accuracy_score(y_true=y_val, y_pred=y_pred)}\\nTest F1-Score : {f1_score(y_true=y_val, y_pred=y_pred, average='weighted')}\")\n",
    "\n",
    "name = f\"logreg_{best_model['C']}_{best_model['acc']:.2f}_{best_model['f1']:.2f}\"\n",
    "pickle.dump(best_model['model'], open(os.path.join(models_path, name), 'wb'))\n",
    "\n",
    "result = pd.DataFrame(ALL_MODELS)\n",
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.8166431497120059\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.roc_curve.RocCurveDisplay at 0x2682e5572e0>"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred = best_model['model'].predict(X_val)\n",
    "print(roc_auc_score(y_true=y_val, y_score=y_pred))\n",
    "RocCurveDisplay.from_predictions(y_true=y_val, y_pred=y_pred, name=\"ROC Curve\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Kernel SVM (Polynomial)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.svm import SVC\n",
    "\n",
    "C = [0.001, 0.01, 0.1, 1, 10]\n",
    "ALL_MODELS = []\n",
    "\n",
    "best_model = {'C' : 0, 'acc' : 0, 'f1' : 0, 'model' : None}\n",
    "for c in C:\n",
    "    model = SVC(C=c, kernel='poly', class_weight='balanced')\n",
    "    scores = cross_val_score(model, X, y, cv=5)\n",
    "    model.fit(X_train, y_train)\n",
    "    y_pred = model.predict(X_val)\n",
    "    acc = accuracy_score(y_true=y_val, y_pred=y_pred)\n",
    "    f1 = f1_score(y_true=y_val, y_pred=y_pred, average='weighted')\n",
    "    print(f\"C : {c:.3f}\\tAccuracy : {acc:.3f}\\tF1 Score : {f1:.3f}\")\n",
    "    ALL_MODELS.append(pd.Series({'C' : c, 'cv_score_fold1' : scores[0], 'cv_score_fold2' : scores[1], 'cv_score_fold3' : scores[2], 'cv_score_fold4' : scores[3], 'cv_score_fold5' : scores[4], 'mean' : scores.mean(), 'std' : scores.std(), 'accuracy' : acc, 'f1_score' : f1}))\n",
    "    if best_model['f1'] < f1:\n",
    "        best_model = {'C' : c, 'acc' : acc, 'f1' : f1, 'model' : copy.deepcopy(model)}\n",
    "\n",
    "y_pred = best_model['model'].predict(X_val)\n",
    "print(f\"Best Model based on Validation Accuracy : {best_model}\\nValidation Accuracy : {accuracy_score(y_true=y_val, y_pred=y_pred)}\\nTest F1-Score : {f1_score(y_true=y_val, y_pred=y_pred, average='weighted')}\")\n",
    "\n",
    "name = f\"svm_poly_{best_model['C']}_{best_model['acc']:.2f}_{best_model['f1']:.2f}\"\n",
    "pickle.dump(best_model['model'], open(os.path.join(models_path, name), 'wb'))\n",
    "\n",
    "result = pd.DataFrame(ALL_MODELS)\n",
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>C</th>\n",
       "      <th>cv_score_fold1</th>\n",
       "      <th>cv_score_fold2</th>\n",
       "      <th>cv_score_fold3</th>\n",
       "      <th>cv_score_fold4</th>\n",
       "      <th>cv_score_fold5</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>accuracy</th>\n",
       "      <th>f1_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.001</td>\n",
       "      <td>0.882008</td>\n",
       "      <td>0.881553</td>\n",
       "      <td>0.879451</td>\n",
       "      <td>0.874253</td>\n",
       "      <td>0.874806</td>\n",
       "      <td>0.878414</td>\n",
       "      <td>0.003292</td>\n",
       "      <td>0.897600</td>\n",
       "      <td>0.887315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.010</td>\n",
       "      <td>0.880902</td>\n",
       "      <td>0.880889</td>\n",
       "      <td>0.877682</td>\n",
       "      <td>0.863747</td>\n",
       "      <td>0.863305</td>\n",
       "      <td>0.873305</td>\n",
       "      <td>0.008072</td>\n",
       "      <td>0.874820</td>\n",
       "      <td>0.884108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.100</td>\n",
       "      <td>0.887980</td>\n",
       "      <td>0.888741</td>\n",
       "      <td>0.879341</td>\n",
       "      <td>0.861646</td>\n",
       "      <td>0.858438</td>\n",
       "      <td>0.875229</td>\n",
       "      <td>0.012872</td>\n",
       "      <td>0.847175</td>\n",
       "      <td>0.865838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.907995</td>\n",
       "      <td>0.891396</td>\n",
       "      <td>0.883765</td>\n",
       "      <td>0.847047</td>\n",
       "      <td>0.849259</td>\n",
       "      <td>0.875892</td>\n",
       "      <td>0.023976</td>\n",
       "      <td>0.838217</td>\n",
       "      <td>0.859652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10.000</td>\n",
       "      <td>0.919053</td>\n",
       "      <td>0.890069</td>\n",
       "      <td>0.886419</td>\n",
       "      <td>0.810551</td>\n",
       "      <td>0.831343</td>\n",
       "      <td>0.867487</td>\n",
       "      <td>0.040189</td>\n",
       "      <td>0.836337</td>\n",
       "      <td>0.858536</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        C  cv_score_fold1  cv_score_fold2  cv_score_fold3  cv_score_fold4  \\\n",
       "0   0.001        0.882008        0.881553        0.879451        0.874253   \n",
       "1   0.010        0.880902        0.880889        0.877682        0.863747   \n",
       "2   0.100        0.887980        0.888741        0.879341        0.861646   \n",
       "3   1.000        0.907995        0.891396        0.883765        0.847047   \n",
       "4  10.000        0.919053        0.890069        0.886419        0.810551   \n",
       "\n",
       "   cv_score_fold5      mean       std  accuracy  f1_score  \n",
       "0        0.874806  0.878414  0.003292  0.897600  0.887315  \n",
       "1        0.863305  0.873305  0.008072  0.874820  0.884108  \n",
       "2        0.858438  0.875229  0.012872  0.847175  0.865838  \n",
       "3        0.849259  0.875892  0.023976  0.838217  0.859652  \n",
       "4        0.831343  0.867487  0.040189  0.836337  0.858536  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "name = f\"svm_poly_{best_model['C']}_{best_model['acc']:.2f}_{best_model['f1']:.2f}\"\n",
    "pickle.dump(best_model['model'], open(os.path.join(models_path, name), 'wb'))\n",
    "\n",
    "result = pd.DataFrame(ALL_MODELS)\n",
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6673107067279163\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.roc_curve.RocCurveDisplay at 0x1b5ddfb8550>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred = best_model['model'].predict(X_val)\n",
    "print(roc_auc_score(y_true=y_val, y_score=y_pred))\n",
    "RocCurveDisplay.from_predictions(y_true=y_val, y_pred=y_pred, name=\"ROC Curve\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Kernel SVM (RBF)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Model based on Validation Accuracy : {'C': 10, 'acc': 0.8355634192192857, 'f1': 0.8576250951366612, 'model': SVC(C=10, class_weight='balanced')}\n",
      "Validation Accuracy : 0.8355634192192857\n",
      "Test F1-Score : 0.8576250951366612\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<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>C</th>\n",
       "      <th>cv_score_fold1</th>\n",
       "      <th>cv_score_fold2</th>\n",
       "      <th>cv_score_fold3</th>\n",
       "      <th>cv_score_fold4</th>\n",
       "      <th>cv_score_fold5</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>accuracy</th>\n",
       "      <th>f1_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.001</td>\n",
       "      <td>0.771757</td>\n",
       "      <td>0.771843</td>\n",
       "      <td>0.737669</td>\n",
       "      <td>0.690445</td>\n",
       "      <td>0.737669</td>\n",
       "      <td>0.741876</td>\n",
       "      <td>0.029905</td>\n",
       "      <td>0.732058</td>\n",
       "      <td>0.779582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.010</td>\n",
       "      <td>0.776180</td>\n",
       "      <td>0.772285</td>\n",
       "      <td>0.762553</td>\n",
       "      <td>0.688564</td>\n",
       "      <td>0.739328</td>\n",
       "      <td>0.747782</td>\n",
       "      <td>0.032259</td>\n",
       "      <td>0.802610</td>\n",
       "      <td>0.833967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.100</td>\n",
       "      <td>0.808471</td>\n",
       "      <td>0.806016</td>\n",
       "      <td>0.844282</td>\n",
       "      <td>0.732250</td>\n",
       "      <td>0.718978</td>\n",
       "      <td>0.781999</td>\n",
       "      <td>0.048173</td>\n",
       "      <td>0.813668</td>\n",
       "      <td>0.842732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.812230</td>\n",
       "      <td>0.796395</td>\n",
       "      <td>0.846715</td>\n",
       "      <td>0.704048</td>\n",
       "      <td>0.647091</td>\n",
       "      <td>0.761296</td>\n",
       "      <td>0.074085</td>\n",
       "      <td>0.820303</td>\n",
       "      <td>0.847588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10.000</td>\n",
       "      <td>0.814332</td>\n",
       "      <td>0.790312</td>\n",
       "      <td>0.829684</td>\n",
       "      <td>0.648640</td>\n",
       "      <td>0.655607</td>\n",
       "      <td>0.747715</td>\n",
       "      <td>0.079083</td>\n",
       "      <td>0.835563</td>\n",
       "      <td>0.857625</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        C  cv_score_fold1  cv_score_fold2  cv_score_fold3  cv_score_fold4  \\\n",
       "0   0.001        0.771757        0.771843        0.737669        0.690445   \n",
       "1   0.010        0.776180        0.772285        0.762553        0.688564   \n",
       "2   0.100        0.808471        0.806016        0.844282        0.732250   \n",
       "3   1.000        0.812230        0.796395        0.846715        0.704048   \n",
       "4  10.000        0.814332        0.790312        0.829684        0.648640   \n",
       "\n",
       "   cv_score_fold5      mean       std  accuracy  f1_score  \n",
       "0        0.737669  0.741876  0.029905  0.732058  0.779582  \n",
       "1        0.739328  0.747782  0.032259  0.802610  0.833967  \n",
       "2        0.718978  0.781999  0.048173  0.813668  0.842732  \n",
       "3        0.647091  0.761296  0.074085  0.820303  0.847588  \n",
       "4        0.655607  0.747715  0.079083  0.835563  0.857625  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "\n",
    "C = [0.001, 0.01, 0.1, 1, 10]\n",
    "ALL_MODELS = []\n",
    "\n",
    "best_model = {'C' : 0, 'acc' : 0, 'f1' : 0, 'model' : None}\n",
    "for c in C:\n",
    "    model = SVC(C=c, kernel='rbf', class_weight='balanced')\n",
    "    scores = cross_val_score(model, X, y, cv=5)\n",
    "    model.fit(X_train, y_train)\n",
    "    y_pred = model.predict(X_val)\n",
    "    acc = accuracy_score(y_true=y_val, y_pred=y_pred)\n",
    "    f1 = f1_score(y_true=y_val, y_pred=y_pred, average='weighted')\n",
    "    #print(f\"C : {c:.3f}\\tAccuracy : {acc:.3f}\\tF1 Score : {f1:.3f}\")\n",
    "    ALL_MODELS.append(pd.Series({'C' : c, 'cv_score_fold1' : scores[0], 'cv_score_fold2' : scores[1], 'cv_score_fold3' : scores[2], 'cv_score_fold4' : scores[3], 'cv_score_fold5' : scores[4], 'mean' : scores.mean(), 'std' : scores.std(), 'accuracy' : acc, 'f1_score' : f1}))\n",
    "    if best_model['f1'] < f1:\n",
    "        best_model = {'C' : c, 'acc' : acc, 'f1' : f1, 'model' : copy.deepcopy(model)}\n",
    "\n",
    "y_pred = best_model['model'].predict(X_val)\n",
    "print(f\"Best Model based on Validation Accuracy : {best_model}\\nValidation Accuracy : {accuracy_score(y_true=y_val, y_pred=y_pred)}\\nTest F1-Score : {f1_score(y_true=y_val, y_pred=y_pred, average='weighted')}\")\n",
    "\n",
    "name = f\"svm_rbf_{best_model['C']}_{best_model['acc']:.2f}_{best_model['f1']:.2f}\"\n",
    "pickle.dump(best_model['model'], open(os.path.join(models_path, name), 'wb'))\n",
    "\n",
    "result = pd.DataFrame(ALL_MODELS)\n",
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC Score : 0.813\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.roc_curve.RocCurveDisplay at 0x168b785c160>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred = best_model['model'].predict(X_val)\n",
    "print(f\"AUC Score : {roc_auc_score(y_true=y_val, y_score=y_pred):.3f}\")\n",
    "RocCurveDisplay.from_predictions(y_true=y_val, y_pred=y_pred, name=\"ROC Curve\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Linear SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n",
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Model based on Validation Accuracy : {'C': 10, 'acc': 0.8553577352648457, 'f1': 0.8707695093143715, 'model': LinearSVC(C=10, class_weight='balanced')}\n",
      "Validation Accuracy : 0.8553577352648457\n",
      "Test F1-Score : 0.8707695093143715\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\ayana\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sklearn\\svm\\_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<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>C</th>\n",
       "      <th>cv_score_fold1</th>\n",
       "      <th>cv_score_fold2</th>\n",
       "      <th>cv_score_fold3</th>\n",
       "      <th>cv_score_fold4</th>\n",
       "      <th>cv_score_fold5</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>accuracy</th>\n",
       "      <th>f1_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.001</td>\n",
       "      <td>0.902134</td>\n",
       "      <td>0.880889</td>\n",
       "      <td>0.888078</td>\n",
       "      <td>0.621986</td>\n",
       "      <td>0.857554</td>\n",
       "      <td>0.830128</td>\n",
       "      <td>0.105067</td>\n",
       "      <td>0.822625</td>\n",
       "      <td>0.848022</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.010</td>\n",
       "      <td>0.904899</td>\n",
       "      <td>0.719199</td>\n",
       "      <td>0.883101</td>\n",
       "      <td>0.190998</td>\n",
       "      <td>0.871157</td>\n",
       "      <td>0.713871</td>\n",
       "      <td>0.269553</td>\n",
       "      <td>0.824284</td>\n",
       "      <td>0.849211</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.100</td>\n",
       "      <td>0.882782</td>\n",
       "      <td>0.884096</td>\n",
       "      <td>0.216434</td>\n",
       "      <td>0.542690</td>\n",
       "      <td>0.885534</td>\n",
       "      <td>0.682307</td>\n",
       "      <td>0.267858</td>\n",
       "      <td>0.824505</td>\n",
       "      <td>0.849401</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.883114</td>\n",
       "      <td>0.765760</td>\n",
       "      <td>0.887525</td>\n",
       "      <td>0.845941</td>\n",
       "      <td>0.882659</td>\n",
       "      <td>0.853000</td>\n",
       "      <td>0.046128</td>\n",
       "      <td>0.823842</td>\n",
       "      <td>0.848854</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10.000</td>\n",
       "      <td>0.875484</td>\n",
       "      <td>0.835213</td>\n",
       "      <td>0.700288</td>\n",
       "      <td>0.847158</td>\n",
       "      <td>0.720969</td>\n",
       "      <td>0.795822</td>\n",
       "      <td>0.071081</td>\n",
       "      <td>0.855358</td>\n",
       "      <td>0.870770</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        C  cv_score_fold1  cv_score_fold2  cv_score_fold3  cv_score_fold4  \\\n",
       "0   0.001        0.902134        0.880889        0.888078        0.621986   \n",
       "1   0.010        0.904899        0.719199        0.883101        0.190998   \n",
       "2   0.100        0.882782        0.884096        0.216434        0.542690   \n",
       "3   1.000        0.883114        0.765760        0.887525        0.845941   \n",
       "4  10.000        0.875484        0.835213        0.700288        0.847158   \n",
       "\n",
       "   cv_score_fold5      mean       std  accuracy  f1_score  \n",
       "0        0.857554  0.830128  0.105067  0.822625  0.848022  \n",
       "1        0.871157  0.713871  0.269553  0.824284  0.849211  \n",
       "2        0.885534  0.682307  0.267858  0.824505  0.849401  \n",
       "3        0.882659  0.853000  0.046128  0.823842  0.848854  \n",
       "4        0.720969  0.795822  0.071081  0.855358  0.870770  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import LinearSVC\n",
    "\n",
    "C = [0.001, 0.01, 0.1, 1, 10]\n",
    "ALL_MODELS = []\n",
    "\n",
    "best_model = {'C' : 0, 'acc' : 0, 'f1' : 0, 'model' : None}\n",
    "for c in C:\n",
    "    model = LinearSVC(C=c, class_weight='balanced')\n",
    "    scores = cross_val_score(model, X, y, cv=5)\n",
    "    model.fit(X_train, y_train)\n",
    "    y_pred = model.predict(X_val)\n",
    "    acc = accuracy_score(y_true=y_val, y_pred=y_pred)\n",
    "    f1 = f1_score(y_true=y_val, y_pred=y_pred, average='weighted')\n",
    "    #print(f\"C : {c:.3f}\\tAccuracy : {acc:.3f}\\tF1 Score : {f1:.3f}\")\n",
    "    ALL_MODELS.append(pd.Series({'C' : c, 'cv_score_fold1' : scores[0], 'cv_score_fold2' : scores[1], 'cv_score_fold3' : scores[2], 'cv_score_fold4' : scores[3], 'cv_score_fold5' : scores[4], 'mean' : scores.mean(), 'std' : scores.std(), 'accuracy' : acc, 'f1_score' : f1}))\n",
    "    if best_model['f1'] < f1:\n",
    "        best_model = {'C' : c, 'acc' : acc, 'f1' : f1, 'model' : copy.deepcopy(model)}\n",
    "\n",
    "y_pred = best_model['model'].predict(X_val)\n",
    "print(f\"Best Model based on Validation Accuracy : {best_model}\\nValidation Accuracy : {accuracy_score(y_true=y_val, y_pred=y_pred)}\\nTest F1-Score : {f1_score(y_true=y_val, y_pred=y_pred, average='weighted')}\")\n",
    "\n",
    "name = f\"svm_linear_{best_model['C']}_{best_model['acc']:.2f}_{best_model['f1']:.2f}\"\n",
    "pickle.dump(best_model['model'], open(os.path.join(models_path, name), 'wb'))\n",
    "\n",
    "result = pd.DataFrame(ALL_MODELS)\n",
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC Score : 0.793\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.roc_curve.RocCurveDisplay at 0x168b7650af0>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred = best_model['model'].predict(X_val)\n",
    "print(f\"AUC Score : {roc_auc_score(y_true=y_val, y_score=y_pred):.3f}\")\n",
    "RocCurveDisplay.from_predictions(y_true=y_val, y_pred=y_pred, name=\"ROC Curve\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Model based on Validation Accuracy : {'depth': 8, 'acc': 0.9016919163994249, 'f1': 0.8985275975658432, 'model': DecisionTreeClassifier(max_depth=8)}\n",
      "Validation Accuracy : 0.9016919163994249\n",
      "Test F1-Score : 0.8985275975658432\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<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>depth</th>\n",
       "      <th>cv_score_fold1</th>\n",
       "      <th>cv_score_fold2</th>\n",
       "      <th>cv_score_fold3</th>\n",
       "      <th>cv_score_fold4</th>\n",
       "      <th>cv_score_fold5</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>accuracy</th>\n",
       "      <th>f1_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.883101</td>\n",
       "      <td>0.882990</td>\n",
       "      <td>0.882990</td>\n",
       "      <td>0.882990</td>\n",
       "      <td>0.883015</td>\n",
       "      <td>0.000043</td>\n",
       "      <td>0.886984</td>\n",
       "      <td>0.833861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.884428</td>\n",
       "      <td>0.870714</td>\n",
       "      <td>0.398142</td>\n",
       "      <td>0.873369</td>\n",
       "      <td>0.781931</td>\n",
       "      <td>0.191968</td>\n",
       "      <td>0.890412</td>\n",
       "      <td>0.887267</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.809113</td>\n",
       "      <td>0.883212</td>\n",
       "      <td>0.396483</td>\n",
       "      <td>0.832780</td>\n",
       "      <td>0.760918</td>\n",
       "      <td>0.184478</td>\n",
       "      <td>0.900476</td>\n",
       "      <td>0.887109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>0.883114</td>\n",
       "      <td>0.800708</td>\n",
       "      <td>0.871489</td>\n",
       "      <td>0.537602</td>\n",
       "      <td>0.809666</td>\n",
       "      <td>0.780516</td>\n",
       "      <td>0.125751</td>\n",
       "      <td>0.900476</td>\n",
       "      <td>0.887261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.802367</td>\n",
       "      <td>0.716656</td>\n",
       "      <td>0.445587</td>\n",
       "      <td>0.721522</td>\n",
       "      <td>0.713827</td>\n",
       "      <td>0.147295</td>\n",
       "      <td>0.901802</td>\n",
       "      <td>0.893772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.829352</td>\n",
       "      <td>0.831785</td>\n",
       "      <td>0.528091</td>\n",
       "      <td>0.719420</td>\n",
       "      <td>0.758330</td>\n",
       "      <td>0.126891</td>\n",
       "      <td>0.900807</td>\n",
       "      <td>0.894109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7.0</td>\n",
       "      <td>0.889528</td>\n",
       "      <td>0.805685</td>\n",
       "      <td>0.838974</td>\n",
       "      <td>0.457863</td>\n",
       "      <td>0.720305</td>\n",
       "      <td>0.742471</td>\n",
       "      <td>0.152596</td>\n",
       "      <td>0.899149</td>\n",
       "      <td>0.890263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8.0</td>\n",
       "      <td>0.881676</td>\n",
       "      <td>0.812984</td>\n",
       "      <td>0.816412</td>\n",
       "      <td>0.474453</td>\n",
       "      <td>0.491705</td>\n",
       "      <td>0.695446</td>\n",
       "      <td>0.175202</td>\n",
       "      <td>0.901692</td>\n",
       "      <td>0.898528</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9.0</td>\n",
       "      <td>0.882119</td>\n",
       "      <td>0.815970</td>\n",
       "      <td>0.816191</td>\n",
       "      <td>0.499558</td>\n",
       "      <td>0.499005</td>\n",
       "      <td>0.702568</td>\n",
       "      <td>0.167726</td>\n",
       "      <td>0.900476</td>\n",
       "      <td>0.895583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10.0</td>\n",
       "      <td>0.884773</td>\n",
       "      <td>0.788542</td>\n",
       "      <td>0.815085</td>\n",
       "      <td>0.479208</td>\n",
       "      <td>0.497346</td>\n",
       "      <td>0.692991</td>\n",
       "      <td>0.170175</td>\n",
       "      <td>0.895720</td>\n",
       "      <td>0.892624</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   depth  cv_score_fold1  cv_score_fold2  cv_score_fold3  cv_score_fold4  \\\n",
       "0    1.0        0.883003        0.883101        0.882990        0.882990   \n",
       "1    2.0        0.883003        0.884428        0.870714        0.398142   \n",
       "2    3.0        0.883003        0.809113        0.883212        0.396483   \n",
       "3    4.0        0.883114        0.800708        0.871489        0.537602   \n",
       "4    5.0        0.883003        0.802367        0.716656        0.445587   \n",
       "5    6.0        0.883003        0.829352        0.831785        0.528091   \n",
       "6    7.0        0.889528        0.805685        0.838974        0.457863   \n",
       "7    8.0        0.881676        0.812984        0.816412        0.474453   \n",
       "8    9.0        0.882119        0.815970        0.816191        0.499558   \n",
       "9   10.0        0.884773        0.788542        0.815085        0.479208   \n",
       "\n",
       "   cv_score_fold5      mean       std  accuracy  f1_score  \n",
       "0        0.882990  0.883015  0.000043  0.886984  0.833861  \n",
       "1        0.873369  0.781931  0.191968  0.890412  0.887267  \n",
       "2        0.832780  0.760918  0.184478  0.900476  0.887109  \n",
       "3        0.809666  0.780516  0.125751  0.900476  0.887261  \n",
       "4        0.721522  0.713827  0.147295  0.901802  0.893772  \n",
       "5        0.719420  0.758330  0.126891  0.900807  0.894109  \n",
       "6        0.720305  0.742471  0.152596  0.899149  0.890263  \n",
       "7        0.491705  0.695446  0.175202  0.901692  0.898528  \n",
       "8        0.499005  0.702568  0.167726  0.900476  0.895583  \n",
       "9        0.497346  0.692991  0.170175  0.895720  0.892624  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "best_model = {'depth' : 0, 'acc' : 0, 'f1' : 0, 'model' : None}\n",
    "ALL_MODELS = []\n",
    "\n",
    "for i in range(1, 11):\n",
    "    model = DecisionTreeClassifier(max_depth=i)\n",
    "    scores = cross_val_score(model, X, y, cv=5)\n",
    "    model.fit(X_train, y_train)\n",
    "    y_pred = model.predict(X_val)\n",
    "    acc = accuracy_score(y_true=y_val, y_pred=y_pred)\n",
    "    f1 = f1_score(y_true=y_val, y_pred=y_pred, average='weighted')\n",
    "    #print(f\"Depth : {c:.3f}\\tAccuracy : {acc:.3f}\\tF1 Score : {f1:.3f}\")\n",
    "    ALL_MODELS.append(pd.Series({'depth' : i, 'cv_score_fold1' : scores[0], 'cv_score_fold2' : scores[1], 'cv_score_fold3' : scores[2], 'cv_score_fold4' : scores[3], 'cv_score_fold5' : scores[4], 'mean' : scores.mean(), 'std' : scores.std(), 'accuracy' : acc, 'f1_score' : f1}))\n",
    "    if best_model['f1'] < f1:\n",
    "        best_model = {'depth' : i, 'acc' : acc, 'f1' : f1, 'model' : copy.deepcopy(model)}\n",
    "\n",
    "y_pred = best_model['model'].predict(X_val)\n",
    "print(f\"Best Model based on Validation Accuracy : {best_model}\\nValidation Accuracy : {accuracy_score(y_true=y_val, y_pred=y_pred)}\\nTest F1-Score : {f1_score(y_true=y_val, y_pred=y_pred, average='weighted')}\")\n",
    "\n",
    "name = f\"dtree_{best_model['depth']}_{best_model['acc']:.2f}_{best_model['f1']:.2f}\"\n",
    "pickle.dump(best_model['model'], open(os.path.join(models_path, name), 'wb'))\n",
    "\n",
    "result = pd.DataFrame(ALL_MODELS)\n",
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC Score : 0.725\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.roc_curve.RocCurveDisplay at 0x168b7c939a0>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEGCAYAAABo25JHAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAAA2o0lEQVR4nO3deXxU9bn48c+ThSTsS1gSIBJ2AiSIEYsLKiiiJCqVinqt194qbdXaurV20Vr13vbW295qa61Lvdpqodb+VBJF3EBcy6Ik7IIQJUyAJGwhIes8vz/OSTqEkDkhmUySed6vV17M2Z8zM5zvnO/3e56vqCrGGGMiV1S4AzDGGBNeVhAYY0yEs4LAGGMinBUExhgT4awgMMaYCBcT7gBaKjExUUeMGBHuMIwxplNZu3ZtiaoObGpZpysIRowYwZo1a8IdhjHGdCoi8sWJllnVkDHGRDgrCIwxJsJZQWCMMRHOCgJjjIlwVhAYY0yEC1lBICJPi8g+EdlwguUiIo+IyHYRyReRqaGKxRhjzImF8o7gGWBOM8svBsa4fwuBx0IYizHGmBMIWUGgqiuB/c2schnwZ3V8DPQVkaRQxWOMMZ3V9n1H+O1bn7F1T1lI9h/OB8qGArsCpgvdeUWNVxSRhTh3DaSkpLRLcMYYE0679leQk+8jJ6+IzUWHEYEBPeMYN6RXmx+rUzxZrKpPAE8AZGZm2kg6xpguac+hSnLzfeTmF7Fu10EApqb05d6sNOamJzG4d3xIjhvOgmA3MDxgepg7zxhjIkbJkSqWri8iJ7+I1QX7UYVJQ3tz98XjmTs5ieH9u4c8hnAWBEuAW0RkMXAGcEhVj6sWMsaYruZQRQ3LNu4hJ9/HB9tL8CuMGdST2y4YS1Z6EiMH9mzXeEJWEIjIIuA8IFFECoGfAbEAqvpH4DXgEmA7UAF8I1SxGGNMuB2pquWtTXvJyfOxclsxNXXKKQO6c9N5o8nOSA5J3b9XISsIVPXqIMsVuDlUxzfGmHA7Wl3H8q37yMnz8c6WfVTV+knuE883zkolOz2ZSUN7IyLhDrNzNBYbY0xnUVVbx3uflZCT7+PNTXupqK4jsWccV09LISs9iakp/YiKCv/FP5AVBMYY00q1dX4+/LyUnDwfyzbu4XBlLX27x3LZlGSy05M5Y+QAojvYxT+QFQTGGHMS/H5lVcF+cvN9LF2/h9LyanrFxXDhxMFkZyRz9uhEYqM7Rzo3KwiMMcYjVWXdroPk5BXx6nofew9XkRAbzawJg8jOSObcsQOJj40Od5gtZgWBMcY0Q1XZVHSYnLwicvN9FB44SrfoKM4bN5DsjGRmTRhE926d+1LauaM3xpgQ2b6vjCV5ReTm+dhRUk5MlHD2mERuu2AsF04cTO/42HCH2GY8FQQiEgVkAMnAUWCDqu4LZWDGGNPevigtJze/iJw8H1v2lCEC00cO4MYZI5kzcQj9enQLd4gh0WxBICKjgB8CFwDbgGIgHhgrIhXA48CzquoPdaDGGBMKvoNHeTXfqfbJKzwEQOYp/bgvO41LJicxKET5fTqSYHcED+KME/At9wGwBiIyCLgG+DrwbGjCM8aYtldcVsVr652L/+qCAwBMHtqHH18ynrnpyQztmxDmCNtXswVBc08Hu1VDv23rgIwxJhQOVlTz+gYnv89Hn5fiVxg3uBd3zh5LVnoyIxJ7hDvEsDnpxmIRuVBV32zLYIwxpi2VVdbwppvf571tJdT6ldTEHtxy/miyMpIZOzh8+X06ktb0GvoTYKPEGGM6lKPVdby9xbn4L99aTHWtn6F9E/jmOU5+n4nJHSO/T0cSrLF4yYkWAQPaPhxjjGm5qto63t1aTE5+EW9vdvL7DOoVxzXTUsjOSGZqSl+7+Dcj2B3BOcC1wJFG8wWYFpKIjDHGg5o6Px9sLyE3v4hlG/dQVllLv+6xXH7qULLTk5mW2r9D5/fpSIIVBB8DFar6buMFIrI1NCEZY0zT6vzKqp37ycn3sXR9EQcqaugVF8NFk4aQnZHMmaMGdJr8Ph1JsF5DFzezbEbbh2OMMcfy+5VPdx1w8/sUUVxWRfdu0VwwwUnuNmNsInExnS+/T0diKSaMMR2OqrLRd5icPGcg990Hj9ItJoqZ45zkbjPHDyKhm13824oVBMaYDuOzvWXk5PnIyfNRUFpBTJQwY+xA7pg9lgvTBtOrC+X36UisIDDGhNXOknJy3V/+W/eWESVw5qhEvn3uKOZMGkLf7l0zv09HYgWBMabd7T54lFfzfeTkFbF+t5Pf5/QR/bj/solcPCmJgb3iwhxhZPFcEIjIfap634mmjTGmOfvKKnktv4ic/CLWfuHk98kY1oefzp3AJZOTSI6w/D4dSUvuCNYGmTbGmGPsL3fz++T5+HhnKaowfkgv7rpoHFnpSZwyIHLz+3QkngsCVc1pbtoYYwAOV9bwxkYnxcMH2538PiMTe3DrzDFkZyQxepDl9+logqWY+B2gJ1quqre2eUTGmE6norqWtzbvIyfPx7tbi6mu8zOsXwI3nDOS7Iwk0pIsv09HFuyOYE27RGGM6XQqa+pYsbWYnHwf72zex9GaOgb3juPar5xCdkYSU4Zbfp/OItiTxccMOCMi3VW1IrQhGWM6qpo6P+9vKyEn38cbG/dypKqW/j26ccVpTn6f00f0J8ry+3Q6Xscsno6TdronkCIiGTijlt0UyuCMMeFX51f+uaPUye+zYQ8HK2roHR/DJZOHkJXu5PeJsfw+nZrXxuLfAhcBSwBUNU9ELNeQMV2U36988uUBcvJ8vLp+DyVHqujRLZoL05z8PueMGUi3GLv4dxUt6TW0q1F9X13bh2OMCRdVZf3uQw35fYoOVRIXE8WsCYPISk/m/HGW36er8loQ7BKRMwEVkVjge8Dm0IVljGkPqspWN79Pbn4RX5RWEBstzBgzkB/OGc8FaYPpGWcJCLo6r5/wt4GHgaGAD1gG3ByqoIwxobWj+Ai5+UXk5PnYtu8I0VHCmaMGcPN5o7lo4hD6dLfkbpHEU0GgqiXAv7V05yIyB6cAiQaeUtVfNlqeAjwL9HXXuVtVX2vpcYwxwe3aX8Gr652L/0bfYUTg9BH9eeDySVw8aQiJPS2/T6Ty2mtoJM4F/Ss4D5h9BNymqjua2SYaeBS4ECgEVovIElXdFLDaT4EXVPUxEUkDXgNGnMyJGGOOt/dwJa/mF5GT7+PTLw8CMGV4X+7JSmPu5CSG9IkPb4CmQ/BaNfRXnIv6PHf6KmARcEYz20wDttcXFiKyGLgMCCwIFOjtvu6DU+1kjGmF0iNVLHXz+6wq2I8qpCX15gdzxpGdnszw/t3DHaLpYLwWBN1V9S8B08+JyF1BthkK7AqYLuT4guM+4A0R+S7QA7igqR2JyEJgIUBKSorHkI2JHIeO1rBso3Px//DzUur8yqiBPfjerDFkpSczelDPcIdoOrBguYb6uy+XisjdwGKcX/ELcKpxWutq4BlV/bX70NpfRGSSqvoDV1LVJ4AnADIzM0+Y+8iYSFJeVctbm/eSk1fEys+c/D7D+yfwrRkjyc5IZvyQXpbiwXgS7I5gLc6Fv/7b9K2AZQr8qJltdwPDA6aHufMCfROYA6CqH4lIPJAI7AsSlzERqbKmjuVb9pGbX8TbW/ZSWeNnSO94rpt+CtkZyaQP62MXf9NiwXINpbZi36uBMSKSilMAXAVc02idL4FZwDMiMgGIB4pbcUxjupzqWj/vby8mJ6+INzbuoby6jsSe3bgyczjZGcmcltLP8vuYVmnJCGWTgDScizUAqvrnE62vqrUicgvOMwfRwNOqulFE7gfWqOoS4A7gSRG5DecO43pVtaofE/Fq6/x8vGM/OXk+Xt+4h0NHa+iTEEt2RjLZGcmckdrf8vuYNuO1++jPgPNwCoLXgIuB94ETFgQA7jMBrzWad2/A603AWS2K2Jguyu9X1nzh5Pd5bX0RpeXV9IyLYXbaYLIykjh7tOX3MaHh9Y5gPpABfKqq3xCRwcBzoQvLmMigquQVOvl9Xs0vYs/hSuJjo5g1fjDZGUmcN24Q8bGW38eElteC4Kiq+kWkVkR64zTmDg+2kTHmeKrK5qIycvJ95Ob72LX/KLHRwrljB/GjS8ZzwYTB9LD8PqYdef22rRGRvsCTOD2JjuA8XWyM8Wj7viPk5vvIyfPxeXE50VHCWaMTuXXmGGZPHEKfBMvvY8LDa66h+gFo/igirwO9VTU/dGEZ0zXs2l9BTr6PnLwiNhc5+X3OSO3Pf5ydypyJQxhg+X1MBxDsgbKpzS1T1U/aPiRjOrc9hyqdX/75ReTtOgjA1JS+3JuVxtz0JAb3tvw+pmMJdkfw62aWKTCzDWMxptMqOVLF0vVF5OQVsfoLJ7/PxOTe3H3xeOZOTrL8PqZDC/ZA2fntFYgxnc3Bimo3v08RH35egl9hzKCe3HbBWLLSkxg50PL7mM7BuiYY0wJHqmp5c5Nz8X9vWzE1dcopA7pz03mjyc5IZtyQXuEO0ZgWs4LAmCCOVtfxzpZ95Ob7eGfLPqpq/ST3iecbZ6WSnZ7MpKG9Lb+P6dSsIDCmCVW1dbz3WQk5+T7e3LSXiuo6EnvGcfW0FLLSk5hq+X1MF+I1xYTgDFU5UlXvd4eYHKKqq0IanTHtqLbOz4eflzbk9ymrrKVv91gum5JMdnoyZ4wcQLRd/E0X5PWO4A+AH6eX0P1AGfAP4PQQxWVMu6jzK6sLnORuSzfsYX95Nb3iYrhw4mCyM5I5e3QisZbczXRxXguCM1R1qoh8CqCqB0SkWwjjMiZkVJVPdx1sSO6293AVCbHRzJowiOyMZM4dO9Dy+5iI4rUgqHEHo1cAERmIc4dgTKegqmz0HSY3v4jcfB+FB47SLTqK88YNJDsjmVkTBtG9mzWZmcjk9Zv/CPASMEhE/hMnG+lPQxaVMW1k294ycvKLyM3zsaOknJgo4ewxiXz/grHMnjiY3vGW38cYr7mGnheRtTijiQlwuapuDmlkxpykL0rLyc0vIifPx5Y9ZYjA9JEDuOGckcyZNIT+PaxW05hAXnsNPQIsVtVHQxyPMSfFd/Aor+YXkZPvI7/wEACZp/Tjvuw0LpmcxCDL72PMCXmtGloL/FRExuFUES1W1TWhC8uY4PaVVbJ0/R5y8nys+eIAAJOH9uHHl4xnbnoyQ/smhDlCYzoHr1VDzwLPikh/4Argv0UkRVXHhDQ6Yxo5UF7N6xv3kJvv46PPS/ErjBvciztnjyUrPZkRiT3CHaIxnU5Lu0mMBsYDpwDWRmDaRVllDW9s3Etuvo/3tpVQ61dSE3twy/mjycpIZuxgy+9jTGt4bSP4FTAP+Bz4G/CAqh4MYVwmwlVU1/LOln3k5PlYvrWY6lo/Q/sm8M1znPw+E5Mtv48xbcXrHcHnwHRVLQllMCayVdXW8e7WYnLyi3hr016O1tQxqFcc10xLITsjmakpfe3ib0wIBBuhbLyqbgFWAylujqEGNkKZaa2aOj8fbC8hJ6+INzY5+X36dY9l3tShZKcnMy21v+X3MSbEgt0R3A4spOmRymyEMnNS6vzKP3eWkpNXxOsbijhQUUOvuBgumjSE7Ixkzhw1wPL7GNOOgo1QttB9ebGqVgYuExHrmG088/uVT3cdICeviFfXF1Fc5uT3uTDNSe42Y2wicTGW38eYcPDaRvAh0Hgg+6bmGdNAVdmw+zC5+T5y84vYffAo3WKimDluEFkZScwcb/l9jOkIgrURDAGGAgkicipOegmA3oCNxm2a9NneMnLyfOTk+SgorSAmSpgxdiB3zB7LhWmD6WX5fYzpUIL9HLsIuB4YBvwmYH4Z8OMQxWQ6oZ0l5eTm+cjJ9/HZ3iNECZw5KpFvnzuKOZOG0Le75fcxpqMK1kZQ/0TxFar6j3aKyXQShQcqGvL7bNh9GIDTR/Tj/ssmcvGkJAb2igtzhMYYL4JVDV2rqs8BI0Tk9sbLVfU3TWxmurB9hyt5dX0RuflFrHXz+2QM68NP507gkslJJFt+H2M6nWBVQ/WJW3qezM5FZA7wMBANPKWqv2xinSuB+3C6o+ap6jUncywTOvvLq1m6oYjcvCI+3lmKKowf0ou7LhpHVnoSpwyw/D7GdGbBqoYed//9eUt37I5o9ihwIVAIrBaRJaq6KWCdMcCPgLPc4S8HtfQ4JjQOHa3hjY17yM0v4v3tJdT5lZGJPbh15hiyM5IYPcjy+xjTVbQk19CDwFHgdSAduM2tNjqRacB2Vd3h7mMxcBmwKWCdG4FHVfUAgKrua/EZmDZTUV3LW5ud/D7vbi2mus7PsH4J3HjOSLIzkkhLsvw+xnRFXjtxz1bVH4jIPKAA+CqwEmiuIBgK7AqYLgTOaLTOWAAR+QCn+ug+VX298Y5EZCHOE86kpKQ0XmxaobKmjhVbi8nJ9/H25r1U1vgZ3DuOa79yCtkZSUwZbvl9jOnqvBYE9evNBf6uqofa6OIQA4wBzsPporpSRCY3zmyqqk8ATwBkZmZqWxw4klXX1uf38fHGpr0cqaqlf49uzD9tGNnpyZw+oj9Rlt/HmIjhtSDIFZEtOFVD3xGRgUBlkG12A8MDpoe58wIVAv9U1Rpgp4h8hlMwrPYYl/Gozq98vKOUnDwfr2/cw8GKGnrHx3DJ5CFkpTv5fWIsv48xEcnrCGV3u+0Eh1S1TkTKcer7m7MaGCMiqTgFwFVA4x5BLwNXA/8nIok4VUU7WhC/aYbfr6z98gC5eT5eXb+HkiNVdO8Wzey0wWSlJ3OO5fcxxuC9sTgWuBaY4VYJvQv8sbltVLVWRG4BluHU/z+tqhtF5H5gjaoucZfNFpFNQB1wl6qWnvTZGFSV/MJDDfl9ig5VEhcTxczxg8jOSOb8cYNI6GYXf2PMv4hq8Cp3EXkKiAWedWd9HahT1RtCGFuTMjMzdc2aNe192A5NVdmyp4zcfB85eUV8ub+C2GhhxpiBZGckc0HaYHrGWXI3YyKZiKxV1cymlnm9OpyuqhkB0++ISF7rQzOt8XnxEXLznBQP2/cdITpKOHPUAG45fzQXTRxCn+6W3M0YE5zXgqBOREap6ucAIjISpyrHtLNd+yvIzS8iJ8/HpqLDiMDpI/rzwOWTuHjSEBJ7Wn4fY0zLeC0I7gKWi8gOnFTUpwDfCFlU5hh7DtXn9/Hx6ZcHAZgyvC/3ZKUxd3ISQ/rYGEHGmJMXtCBwu4oewnlSuD4FxFZVrQplYJGu9EgVSzfsISfPx6qC/ahCWlJvfjBnHNnpyQzvb8NBGGPaRrDsozcA/wV8DqQCC93ePiYEDh2tYdlG5+L/4eel1PmVUQN78L1ZY8hKT2b0oJPK/WeMMc0KdkfwfWCiqha77QLPA1YQtKHyqlre2rzXye/zWTE1dcrw/gl8a8ZIsjOSGT+kl6V4MMaEVLCCoFpViwFUdYeIWEtkG6isqWP5ln3k5Pt4Z8s+Kmv8DOkdz79PH0F2RjLpw/rYxd8Y026CFQTDROSRE02r6q2hCavrqa718962YnLyfLy5aS/l1XUk9uzGlZnDyUpPJvOUfpbfxxgTFsEKgrsaTa8NVSBdUW2dn4/q8/ts2MPhylr6JMSSnZFMVnoyXxnZ3/L7GGPCzsuYxaYF/H5ldcF+cvOLeG19EaXl1fSMi3Hy+2QkcfbogXSLsYu/MabjCNZr6EngYVXd0MSyHsACoEpVnw9RfJ2CqpJXeIicPB+v5hex53Al8bFRzBo/mOyMJM4bN4j4WMvvY4zpmIJVDT0K3Csik4ENQDEQj5MqujfwNE5PooijqmwuKiMn30duvo9d+48SGy2cO3YQP7pkPBdMGEwPy+9jjOkEglUNrQOuFJGeQCaQhDMmwWZV3Rr68Dqe7fuOkJPnIyffx47icqKjhLNGJ3LrzDHMnjiEPgmW38cY07l4HY/gCLAitKF0XLv2V7Akz0nrvNnN73NGan++eXYqcyYOYYDl9zHGdGJWdxHEZ3vLmPPblfgVpqb05d6sNOamJzG4t+X3McZ0DVYQBLHRdwi/wovfnk7miP7hDscYY9pci/oxikjEZTrbWVJBlMDkYX3CHYoxxoSEp4JARM50h5Pc4k5niMgfQhpZB1FQUk5y3wQb29cY02V5vSP4X+AioBRAVfOAGaEKqiMpKC0nNbFHuMMwxpiQ8Vw1pKq7Gs3q8iOUqSo7S6wgMMZ0bV4bi3eJyJmAikgs8D1gc+jC6hj2l1dTVlnLiAFWEBhjui6vdwTfBm4GhgK7gSnATSGKqcMoKC0HsDsCY0yX5vWOYJyq/lvgDBE5C/ig7UPqOHYUOwXBCCsIjDFdmNc7gt95nNelFJQ6KSSG9UsIdyjGGBMywbKPTgfOBAaKyO0Bi3oDXb4/ZUFJBcP7JRBrYwYYY7qwYFVD3YCe7nq9AuYfBuaHKqiOYmdJuVULGWO6vGDZR98F3hWRZ1T1i3aKqUNQVQpKy5mWamkljDFdm9fG4goReQiYiDMeAQCqOjMkUXUAxWVVVFTXWY8hY0yX57Xy+3mc9BKpwM+BAmB1iGLqEHaWWI8hY0xk8FoQDFDVPwE1qvquqv4H0GXvBiDgGQJ7mMwY08V5rRqqcf8tEpG5gA/o0pXnO0sqiI0WkvvauAPGmK7N6x3BgyLSB7gDuBN4Cvh+sI1EZI6IbBWR7SJydzPrXSEiKiKZHuMJuYKScob3706MdR01xnRxXoeqzHVfHgLOh4Yni09IRKKBR4ELgUJgtYgsUdVNjdbrhZO76J8tCz20CkrLGWntA8aYCNDsz10RiRaRq0XkThGZ5M7LEpEPgd8H2fc0YLuq7lDVamAxcFkT6z0A/DdQ2fLwQ8Pvd7qOWrI5Y0wkCFbv8SfgBmAA8IiIPAf8D/ArVT01yLZDgcDU1YXuvAYiMhUYrqqvNrcjEVkoImtEZE1xcXGQw7be3rJKKmv81mPIGBMRglUNZQLpquoXkXhgDzBKVUtbe2ARiQJ+A1wfbF1VfQJ4AiAzM1Nbe+xg6ruO2jMExphIEOyOoFpV/QCqWgnsaEEhsBsYHjA9zJ1XrxcwCVghIgXAV4AlHaHB2J4hMMZEkmB3BONFJN99LcAod1oAVdX0ZrZdDYwRkVScAuAq4Jr6hap6CEisnxaRFcCdqrqmxWfRxgpKyomLiSKpt3UdNcZ0fcEKggknu2NVrRWRW4BlOJlKn1bVjSJyP7BGVZec7L5DbWdJBacM6E5UlIQ7FGOMCblgSedalWhOVV8DXms0794TrHtea47VlqzrqDEmktjTUo3U+ZUvSyusodgYEzGsIGjEd/Ao1XXWddQYEzk8FwQikiAi40IZTEdQn2zOHiYzxkQKTwWBiGQD64DX3ekpItJhG3tbo8CeITDGRBivdwT34aSMOAigqutwxibocnaWVJAQG83g3nHhDsUYY9qF14Kgxu33HyjkT/iGQ0FpOacM6I6IdR01xkQGrwXBRhG5BogWkTEi8jvgwxDGFTYFJeVWLWSMiSheC4Lv4oxXXAX8FScd9fdDFFPY1Nb5+XK/dR01xkQWryOUjVfVnwA/CWUw4bb74FFq/WpdR40xEcXrHcGvRWSziDxQPy5BV2RZR40xkchTQaCq5+OMTFYMPC4i60XkpyGNLAwaso7aMwTGmAji+YEyVd2jqo8A38Z5pqDJnEGdWUFJOT3jYkjs2S3coRhjTLvx+kDZBBG5T0TWA/U9hoaFNLIw2FlawYhE6zpqjIksXhuLnwb+Blykqr4QxhNWBSXlpA/rE+4wjDGmXXkqCFR1eqgDCbfqWj+FByq4bEpyuEMxxph21WxBICIvqOqVbpVQ4JPEXkYo61R2HajAr9ZQbIyJPMHuCL7n/psV6kDCrcDGKTbGRKhmG4tVtch9eZOqfhH4B9wU+vDajz1DYIyJVF67j17YxLyL2zKQcCsoLad3fAz9useGOxRjjGlXwdoIvoPzy3+kiOQHLOoFfBDKwNpbQYmTY8i6jhpjIk2wNoK/AkuBXwB3B8wvU9X9IYsqDHaWlJM5ol+4wzDGmHYXrGpIVbUAuBkoC/hDRPqHNrT2U1lTh+/QUWsfMMZEJC93BFnAWpzuo4H1JgqMDFFc7WrX/gpUraHYGBOZmi0IVDXL/bdLDktZz5LNGWMimddcQ2eJSA/39bUi8hsRSQltaO1npz1DYIyJYF67jz4GVIhIBnAH8Dnwl5BF1c4KSsvp36MbfRKs66gxJvJ4LQhqVVWBy4Dfq+qjOF1Iu4SdJeWMGNA93GEYY0xYeC0IykTkR8DXgVdFJAroMj+fC0oqrFrIGBOxvBYEC3AGrv8PVd2DMxbBQyGLqh0dra5jz+FKUq2h2BgTobwOVbkHeB7oIyJZQKWq/jmkkbWTglJrKDbGRDavvYauBFYBXwOuBP4pIvM9bDdHRLaKyHYRubuJ5beLyCYRyReRt0XklJaeQGsVWLI5Y0yE8zpC2U+A01V1H4CIDATeAl480QYiEg08ipOwrhBYLSJLVHVTwGqfApmqWuHmNfoVTjVUu9lpdwTGmAjntY0gqr4QcJV62HYasF1Vd6hqNbAYp9dRA1VdrqoV7uTHhGEc5IKSchJ7xtEzzmuZaIwxXYvXq9/rIrIMWOROLwBeC7LNUGBXwHQhcEYz638TJ8HdcURkIbAQICWlbZ9jc7KOWtdRY0zk8tpYfBfwOJDu/j2hqj9sqyBE5FogkxP0RFLVJ1Q1U1UzBw4c2FaHBZyqIUstYYyJZMHGIxgD/A8wClgP3Kmquz3uezcwPGB6mDuv8TEuwGmDOFdVqzzuu00cqaqluKyK1IFWEBhjIlewO4KngVzgCpwMpL9rwb5XA2NEJFVEugFXAUsCVxCRU3HuNC5t1AbRLhp6DNkdgTEmggVrI+ilqk+6r7eKyCded6yqtSJyC7AMiAaeVtWNInI/sEZVl+BUBfUE/u6ODPalql7a4rM4SfYMgTHGBC8I4t1f7fXjECQETqtqswWDqr5Go0ZlVb034PUFLY64De0stvTTxhgTrCAoAn4TML0nYFqBmaEIqr3sLC1nSO94ErpFhzsUY4wJm2AD05zfXoGEQ0FJOSOs66gxJsJ5faCsSyoorbDUEsaYiBexBcGhozXsL6+29gFjTMSL2IKgwIanNMYYwHv2UXHHKr7XnU4RkWmhDS206ruOWtWQMSbSeb0j+AMwHbjanS7DySzaae0sKUcEUvpbY7ExJrJ5TTp3hqpOFZFPAVT1gPu0cKdVUFJOcp8E4mOt66gxJrJ5vSOocccXUGgYj8Afsqjawc7SCus6aowxeC8IHgFeAgaJyH8C7wP/FbKo2kFBiWUdNcYY8Fg1pKrPi8haYBZOeonLVXVzSCMLoQPl1Rw6WmMNxcYYg8eCQERSgAogJ3Ceqn4ZqsBCaaf1GDLGmAZeG4tfxWkfECAeSAW2AhNDFFdI2TMExhjzL16rhiYHTovIVOCmkETUDnaWlBMlMLyfNRYbY8xJPVnspp9ubvzhDm1nSTnD+nWnW0zEPlhtjDENvLYR3B4wGQVMBXwhiagdFJSWW7WQMca4vP4k7hXwF4fTZnBZqIIKJVWloKSC1AFWLWSMMeDhjsB9kKyXqt7ZDvGEXMmRao5U1dodgTHGuJq9IxCRGFWtA85qp3hCzsYpNsaYYwW7I1iF0x6wTkSWAH8HyusXqur/C2FsIbHT7Tqaak8VG2MM4P05gnigFGeM4vrnCRTodAVBQUk5MVHCsH4J4Q7FGGM6hGAFwSC3x9AG/lUA1NOQRRVCBaXlDO/fnZho6zpqQqumpobCwkIqKyvDHYqJIPHx8QwbNozY2FjP2wQrCKKBnhxbANTrlAXBzpIKRliPIdMOCgsL6dWrFyNGjECkqf9CxrQtVaW0tJTCwkJSU1M9bxesIChS1ftbF1rHoap8UVrOV0b2D3coJgJUVlZaIWDalYgwYMAAiouLW7RdsPqRLvUN3ldWRUV1nSWbM+3GCgHT3k7mOxesIJh1cqF0TA09hqwgMMaYBs0WBKq6v70CaQ/1BYENSGMiRXR0NFOmTGHSpElkZ2dz8ODBhmUbN25k5syZjBs3jjFjxvDAAw+g+q+mv6VLl5KZmUlaWhqnnnoqd9xxR5PH8LpeqKgqM2fO5PDhww3zXn75ZUSELVu2NMxbsWIFWVlZx2x7/fXX8+KLLwJO4/7dd9/NmDFjmDp1KtOnT2fp0qWtju8Xv/gFo0ePZty4cSxbtqzJdc455xymTJnClClTSE5O5vLLLwfglVdeIT09nSlTppCZmcn7778PQHFxMXPmzGl1bPUiqutMQUk53aKjSO5rXUdNZEhISGDdunVs2LCB/v378+ijjwJw9OhRLr30Uu6++262bt1KXl4eH374IX/4wx8A2LBhA7fccgvPPfccmzZtYs2aNYwePfq4/Xtd70Rqa2tbfY6vvfYaGRkZ9O7du2HeokWLOPvss1m0aJHn/dxzzz0UFRWxYcMGPvnkE15++WXKyspaFdumTZtYvHgxGzdu5PXXX+emm26irq7uuPXee+891q1bx7p165g+fTpf/epXAZg1axZ5eXmsW7eOp59+mhtuuAGAgQMHkpSUxAcffNCq+Op5fY6gS9hZUk7KgO5ER1m9rWlfP8/ZyCbf4eArtkBacm9+lu19SJDp06eTn58PwF//+lfOOussZs+eDUD37t35/e9/z3nnncfNN9/Mr371K37yk58wfvx4wLmz+M53vnPcPptb7/rrrycrK4v58+cD0LNnT44cOcKKFSu455576NevH1u2bOGrX/0qw4cP5+abbwbgvvvuo2fPntx555089NBDvPDCC1RVVTFv3jx+/vOfHxfD888/z8KFCxumjxw5wvvvv8/y5cvJzs5ucpvGKioqePLJJ9m5cydxcXEADB48mCuvvNLbm3sCr7zyCldddRVxcXGkpqYyevRoVq1axfTp05tc//Dhw7zzzjv83//9H+C8Z/XKy8uPqf+//PLLef755znrrNYnfoisO4JSG6fYRKa6ujrefvttLr30UsCpFjrttNOOWWfUqFEcOXKEw4cPs2HDhuOWN8Xreo198sknPPzww3z22WcsWLCAF154oWHZCy+8wIIFC3jjjTfYtm0bq1atYt26daxdu5aVK1cet68PPvjgmBheeeUV5syZw9ixYxkwYABr164NGs/27dtJSUk55q7iRG677baGapzAv1/+8pfHrbt7926GDx/eMD1s2DB27959wn2//PLLzJo165g4XnrpJcaPH8/cuXN5+umnG+ZnZmby3nvvBY3Xi4i5I/D7lS9KKzh37MBwh2IiUEt+ubelo0ePMmXKFHbv3s2ECRO48MILwxJHY9OmTWvo537qqaeyb98+fD4fxcXF9OvXj+HDh/Pwww/zxhtvcOqppwLOL/1t27YxY8aMY/a1f/9+evXq1TC9aNEivve97wFw1VVXsWjRIk477bQT9qZpaS+b//3f/23R+i2xaNGihuqfevPmzWPevHmsXLmSe+65h7feeguAQYMG4fO1zWgAIS0IRGQO8DDOg2lPqeovGy2PA/4MnIaTwmKBqhaEIpaiw5VU1fot2ZyJKPVtBBUVFVx00UU8+uij3HrrraSlpR3363rHjh307NmT3r17M3HiRNauXUtGRkaz+29uvZiYGPx+PwB+v5/q6uqGZT16HPv/8Gtf+xovvvgie/bsYcGCBYDTCPyjH/2Ib33rW83GUH+cqKgo9u/fzzvvvMP69esREerq6hARHnroIQYMGMCBAweO2Xb//v0kJiYyevRovvzySw4fPhz0ruC2225j+fLlx82/6qqruPvuu4+ZN3ToUHbt2tUwXVhYyNChQ5vcb0lJCatWreKll15qcvmMGTPYsWMHJSUlJCYmUllZSUJCG7V3qmpI/nAu/p8DI4FuQB6Q1midm4A/uq+vAv4WbL+nnXaanoz3txXrKT/M1Q+2FZ/U9sa01KZNm8Idgvbo0aPh9SeffKIpKSlaU1OjFRUVmpqaqm+++aaqqlZUVOjcuXP1kUceUVXVvLw8HTVqlG7dulVVVevq6vSxxx47bv/NrffAAw/oD37wA1VVfemll9S53KguX75c586de8x+NmzYoNOnT9cxY8aoz+dTVdVly5bptGnTtKysTFVVCwsLde/evcfFcMYZZ+i2bdtUVfXxxx/XhQsXHrN8xowZ+u6772plZaWOGDGi4XMpKCjQlJQUPXjwoKqq3nXXXXr99ddrVVWVqqru27dPX3jhhWBvcbM2bNig6enpWllZqTt27NDU1FStra1tct3HHntMr7vuumPmbdu2Tf1+v6qqrl27VpOTkxum16xZoxdddFGT+2rquwes0RNcV0PZRjAN2K6qO1S1GljM8YPZXAY8675+EZglIXoCZ6cNWG8i3Kmnnkp6ejqLFi0iISGBV155hQcffJBx48YxefJkTj/9dG655RYA0tPT+e1vf8vVV1/NhAkTmDRpEjt27Dhun82td+ONN/Luu++SkZHBRx99dNxdQKCJEydSVlbG0KFDSUpKAmD27Nlcc801TJ8+ncmTJzN//vwme/HMnTuXFStWAE7Vyrx5845ZfsUVV7Bo0SLi4uJ47rnn+MY3vsGUKVOYP38+Tz31FH369AHgwQcfZODAgaSlpTFp0iSysrI8tRk0Z+LEiVx55ZWkpaUxZ84cHn30UaKjowG45JJLjqnaWbx4MVdfffUx2//jH/9g0qRJTJkyhZtvvpm//e1vDVVZy5cvZ+7cua2Kr56ohiZlkIjMB+ao6g3u9NeBM1T1loB1NrjrFLrTn7vrlDTa10JgIUBKSsppX3zxRYvjeWPjHv6+tpDHrz2NKOs1ZNrB5s2bmTBhQrjD6PKKioq47rrrePPNN8MdSruaMWMGr7zyCv369TtuWVPfPRFZq6qZTe2rU/QaUtUnVDVTVTMHDjy5xt7ZE4fw5HWZVggY08UkJSVx4403HvNAWVdXXFzM7bff3mQhcDJC2Vi8GxgeMD3MndfUOoUiEgP0wWk0NsYYz1rb37+zGThwYMPTx20hlHcEq4ExIpIqIt1wGoOXNFpnCfDv7uv5wDsaqroqY8LAvs6mvZ3Mdy5kBYGq1gK3AMuAzcALqrpRRO4XkUvd1f4EDBCR7cDtwN1N782Yzic+Pp7S0lIrDEy7UXc8gvj4+BZtF7LG4lDJzMzUNWvWhDsMY4KyEcpMOJxohLLmGosj5sliY9pbbGxsi0aJMiZcOkWvIWOMMaFjBYExxkQ4KwiMMSbCdbrGYhEpBlr+aLEjESgJulbXYuccGeycI0NrzvkUVW3yidxOVxC0hoisOVGreVdl5xwZ7JwjQ6jO2aqGjDEmwllBYIwxES7SCoInwh1AGNg5RwY758gQknOOqDYCY4wxx4u0OwJjjDGNWEFgjDERrksWBCIyR0S2ish2ETkuo6mIxInI39zl/xSREWEIs015OOfbRWSTiOSLyNsicko44mxLwc45YL0rRERFpNN3NfRyziJypftZbxSRv7Z3jG3Nw3c7RUSWi8in7vf7knDE2VZE5GkR2eeO4NjUchGRR9z3I19Eprb6oCcazLiz/gHRwOfASKAbkAekNVrnJuCP7uurgL+FO+52OOfzge7u6+9Ewjm76/UCVgIfA5nhjrsdPucxwKdAP3d6ULjjbodzfgL4jvs6DSgId9ytPOcZwFRgwwmWXwIsBQT4CvDP1h6zK94RTAO2q+oOVa0GFgOXNVrnMuBZ9/WLwCypHxG6cwp6zqq6XFUr3MmPcUaM68y8fM4ADwD/DXSFXNBezvlG4FFVPQCgqvvaOca25uWcFagfZb4P4KMTU9WVwP5mVrkM+LM6Pgb6ikhSa47ZFQuCocCugOlCd16T66gzgM4hYEC7RBcaXs450DdxflF0ZkHP2b1lHq6qr7ZnYCHk5XMeC4wVkQ9E5GMRmdNu0YWGl3O+D7hWRAqB14Dvtk9oYdPS/+9B2XgEEUZErgUygXPDHUsoiUgU8Bvg+jCH0t5icKqHzsO561spIpNV9WA4gwqxq4FnVPXXIjId+IuITFJVf7gD6yy64h3BbmB4wPQwd16T64hIDM7tZGm7RBcaXs4ZEbkA+AlwqapWtVNsoRLsnHsBk4AVIlKAU5e6pJM3GHv5nAuBJapao6o7gc9wCobOyss5fxN4AUBVPwLicZKzdVWe/r+3RFcsCFYDY0QkVUS64TQGL2m0zhLg393X84F31G2F6aSCnrOInAo8jlMIdPZ6Ywhyzqp6SFUTVXWEqo7AaRe5VFU78zinXr7bL+PcDSAiiThVRTvaMca25uWcvwRmAYjIBJyCoLhdo2xfS4Dr3N5DXwEOqWpRa3bY5aqGVLVWRG4BluH0OHhaVTeKyP3AGlVdAvwJ5/ZxO06jzFXhi7j1PJ7zQ0BP4O9uu/iXqnpp2IJuJY/n3KV4POdlwGwR2QTUAXepaqe92/V4zncAT4rIbTgNx9d35h92IrIIpzBPdNs9fgbEAqjqH3HaQS4BtgMVwDdafcxO/H4ZY4xpA12xasgYY0wLWEFgjDERzgoCY4yJcFYQGGNMhLOCwBhjIpwVBBFAROpEZF3A34hm1j3SBsd7RkR2usf6xH3as6X7eEpE0tzXP2607MPWxujup/592SAiOSLSN8j6U04ms6WIJIlIrvv6PBE55B53s4j87CT2d2l9Fk4Rubz+fXKn73cfHGwV9zOcH2SdFS15QM8991wP6zWZfVNE/kdEZno9nvHOCoLIcFRVpwT8FbTDMe9S1SnA3TgPsrWIqt6gqpvcyR83WnZm68MD/vW+TMJ5nuTmIOtPwem/3VK3A08GTL/nvjeZODlyWpRGWFWXqOov3cnLcTJu1i+7V1XfOokYO5JngKZyJP0O5/tk2pgVBBFIRHqKMybBJyKyXkSOy9rp/opdGfCL+Rx3/mwR+cjd9u8i0jPI4VYCo91tb3f3tUFEvu/O6yEir4pInjt/gTt/hYhkisgvgQQ3jufdZUfcfxeLyNyAmJ8RkfkiEi0iD4nIanHytX/Lw9vyEW7iLhGZ5p7jpyLyoYiMc59qvR9Y4MaywI39aRFZ5a7bVPZTgCuA1xvPVNVyYC0w2r3b+NiN9yUR6efGcqv8axyJxe6860Xk9yJyJnAp8JAb06iA92COiPw94L1p+DXe0s9QRO5138sNIvKEyDGZer8e8B2Z5q7v9X1p0omyb6rqF8AAERnSkv0ZD8KRb9v+2vcP5wnTde7fSzhPlPd2lyXiPKFY/3DhEfffO4CfuK+jcXL3JOJc2Hu4838I3NvE8Z4B5ruvvwb8EzgNWA/0wHnCeSNwKs5F8smAbfu4/67AHT+gPqaAdepjnAc8677uhpORMQFYCPzUnR8HrAFSm4jzSMD5/R2Y4073BmLc1xcA/3BfXw/8PmD7/wKudV/3xcnr06PRMVKBtQHT5wG57usBQAEwEcgHznXn3w/81n3tA+Lqj9E4jsD3OnDa/Yy/DPisHgOuPcnPsH/A/L8A2QGf0ZPu6xm4+fNP9L40OvdM4KlmvrMjaCIfP86d1RXh/j/V1f66XIoJ06Sj6lRFACAiscB/icgMwI/zS3gwsCdgm9XA0+66L6vqOhE5F6ca4gP3R2E3nF/STXlIRH6Kk/Plmzi5YF5S51cwIvL/gHNwfin/WkT+G+ci8V4Lzmsp8LCIxOFUJaxU1aMiMhtID6jj7oOTeG1no+0TRGSde/6bgTcD1n9WRMbgpCyIPcHxZwOXisid7nQ8kOLuq14Sx+e9OUdEPsV573+Jkyiur6q+6y5/FqdgAqeAeF5EXsbJI+SJOqkZXgeyReRFYC7wA5yss14/w3rni8gPgO5Af5xCPMddtsg93koR6S1OO8uJ3pfA+NYAN3g9nwD7gOST2M40wwqCyPRvwEDgNFWtESc7Z3zgCu5/7Bk4F5BnROQ3wAHgTVW92sMx7lLVF+snRGRWUyup6mduHfklwIMi8raq3u/lJFS1UkRWABcBC3AGLQFn5KbvquqyILs4qqpTRKQ7Ti6bm4FHcAazWa6q88RpWF9xgu0F59fp1uaOQaP3FqeNIKthJyJ9mtl+Ls6v7WzgJyIyuZl1G1sM3IJTzbJGVcvcah2vnyEiEg/8AefubJeI3Mex59M4R41ygvdFRAa3IPYTicd5T00bsjaCyNQH2OcWAucDx41fLM6YxntV9UngKZyh8z4GzhKR+jr/HiIy1uMx3wMuF5HuItIDp1rnPRFJBipU9TmcxHhNNZzWuHcmTfkbTtKt+rsLcC7q36nfRkTGusdskjojt90K3CH/Skten9b3+oBVy3CqyOotA75bX2cuTobXxj7DqeY4IVU9BBwQtx0G+DrwrjhjKgxX1eU4VTh9cKrVAjWOKdC7OO/njfyrkGzpZ1h/0S9x2xIa9ySqb9M5GycL5iG8vS8nayzQ5Fi+5uRZQRCZngcyRWQ9cB2wpYl1zgPy3CqMBcDDqlqMc2FcJCL5OFUK470cUFU/wal3XoXTZvCUqn4KTAZWuVU0PwMebGLzJ4B8cRuLG3kDp7rjLXWGMgSn4NoEfCJOF8THCXL368aSjzPIya+AX7jnHrjdciCtvrEY584h1o1tozvdeL/lwOf1F95m/DtOdVo+Tu+k+3HaLp5zP6dPgUf0+AFmFgN3uY2yoxoduw7IBS52/6Wln6F7vCdxLr7LcKoMA1W679MfcaoAwcP7Ik5HgKeaOqY42Tc/AsaJSKGIfNOdH4vT8aAzpxLvkCz7qDEhJiLzcKrhfhruWDoz932cqqr3hDuWrsbaCIwJMVV9SUQ685jYHUUM8OtwB9EV2R2BMcZEOGsjMMaYCGcFgTHGRDgrCIwxJsJZQWCMMRHOCgJjjIlw/x+j9w+jMjgMxgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred = best_model['model'].predict(X_val)\n",
    "print(f\"AUC Score : {roc_auc_score(y_true=y_val, y_score=y_pred):.3f}\")\n",
    "RocCurveDisplay.from_predictions(y_true=y_val, y_pred=y_pred, name=\"ROC Curve\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Model based on Mean Validation Accuracy : {'depth': 10, 'n_estimators': 500, 'acc': 0.8957204467543957, 'f1': 0.8977363003560574, 'model': RandomForestClassifier(max_depth=10, n_estimators=500)}\n",
      "Test Accuracy : 0.9092115448413137\n",
      "Test F1-Score : 0.8977363003560574\n"
     ]
    },
    {
     "ename": "KeyError",
     "evalue": "'n_esimator'",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
      "\u001b[1;32md:\\IIT JODHPUR\\Semester-3\\ML for Business\\Course Project\\bank_marketing.ipynb Cell 54\u001b[0m in \u001b[0;36m<cell line: 23>\u001b[1;34m()\u001b[0m\n\u001b[0;32m     <a href='vscode-notebook-cell:/d%3A/IIT%20JODHPUR/Semester-3/ML%20for%20Business/Course%20Project/bank_marketing.ipynb#Y100sZmlsZQ%3D%3D?line=19'>20</a>\u001b[0m y_pred \u001b[39m=\u001b[39m best_model[\u001b[39m'\u001b[39m\u001b[39mmodel\u001b[39m\u001b[39m'\u001b[39m]\u001b[39m.\u001b[39mpredict(X_val)\n\u001b[0;32m     <a href='vscode-notebook-cell:/d%3A/IIT%20JODHPUR/Semester-3/ML%20for%20Business/Course%20Project/bank_marketing.ipynb#Y100sZmlsZQ%3D%3D?line=20'>21</a>\u001b[0m \u001b[39mprint\u001b[39m(\u001b[39mf\u001b[39m\u001b[39m\"\u001b[39m\u001b[39mBest Model based on Mean Validation Accuracy : \u001b[39m\u001b[39m{\u001b[39;00mbest_model\u001b[39m}\u001b[39;00m\u001b[39m\\n\u001b[39;00m\u001b[39mTest Accuracy : \u001b[39m\u001b[39m{\u001b[39;00maccuracy_score(y_true\u001b[39m=\u001b[39my_val, y_pred\u001b[39m=\u001b[39my_pred)\u001b[39m}\u001b[39;00m\u001b[39m\\n\u001b[39;00m\u001b[39mTest F1-Score : \u001b[39m\u001b[39m{\u001b[39;00mf1_score(y_true\u001b[39m=\u001b[39my_val, y_pred\u001b[39m=\u001b[39my_pred, average\u001b[39m=\u001b[39m\u001b[39m'\u001b[39m\u001b[39mweighted\u001b[39m\u001b[39m'\u001b[39m)\u001b[39m}\u001b[39;00m\u001b[39m\"\u001b[39m)\n\u001b[1;32m---> <a href='vscode-notebook-cell:/d%3A/IIT%20JODHPUR/Semester-3/ML%20for%20Business/Course%20Project/bank_marketing.ipynb#Y100sZmlsZQ%3D%3D?line=22'>23</a>\u001b[0m name \u001b[39m=\u001b[39m \u001b[39mf\u001b[39m\u001b[39m\"\u001b[39m\u001b[39mrdf_\u001b[39m\u001b[39m{\u001b[39;00mbest_model[\u001b[39m'\u001b[39m\u001b[39mdepth\u001b[39m\u001b[39m'\u001b[39m]\u001b[39m}\u001b[39;00m\u001b[39m_\u001b[39m\u001b[39m{\u001b[39;00mbest_model[\u001b[39m'\u001b[39m\u001b[39mn_esimator\u001b[39m\u001b[39m'\u001b[39m]\u001b[39m}\u001b[39;00m\u001b[39m_\u001b[39m\u001b[39m{\u001b[39;00mbest_model[\u001b[39m'\u001b[39m\u001b[39macc\u001b[39m\u001b[39m'\u001b[39m]\u001b[39m:\u001b[39;00m\u001b[39m.2f\u001b[39m\u001b[39m}\u001b[39;00m\u001b[39m_\u001b[39m\u001b[39m{\u001b[39;00mbest_model[\u001b[39m'\u001b[39m\u001b[39mf1\u001b[39m\u001b[39m'\u001b[39m]\u001b[39m:\u001b[39;00m\u001b[39m.2f\u001b[39m\u001b[39m}\u001b[39;00m\u001b[39m\"\u001b[39m\n\u001b[0;32m     <a href='vscode-notebook-cell:/d%3A/IIT%20JODHPUR/Semester-3/ML%20for%20Business/Course%20Project/bank_marketing.ipynb#Y100sZmlsZQ%3D%3D?line=23'>24</a>\u001b[0m pickle\u001b[39m.\u001b[39mdump(best_model[\u001b[39m'\u001b[39m\u001b[39mmodel\u001b[39m\u001b[39m'\u001b[39m], \u001b[39mopen\u001b[39m(os\u001b[39m.\u001b[39mpath\u001b[39m.\u001b[39mjoin(models_path, name), \u001b[39m'\u001b[39m\u001b[39mwb\u001b[39m\u001b[39m'\u001b[39m))\n\u001b[0;32m     <a href='vscode-notebook-cell:/d%3A/IIT%20JODHPUR/Semester-3/ML%20for%20Business/Course%20Project/bank_marketing.ipynb#Y100sZmlsZQ%3D%3D?line=25'>26</a>\u001b[0m result \u001b[39m=\u001b[39m pd\u001b[39m.\u001b[39mDataFrame(ALL_MODELS)\n",
      "\u001b[1;31mKeyError\u001b[0m: 'n_esimator'"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "n_estimators = [i for i in range(100, 501, 50)]\n",
    "max_depth = [i for i in range(1, 11)]\n",
    "ALL_MODELS = []\n",
    "\n",
    "best_model = {'max_depth' : 0, 'n_estimator' : 0, 'model' : None, 'acc' : 0, 'f1' : 0}\n",
    "for depth in max_depth:\n",
    "    for estimator in n_estimators:\n",
    "        model = RandomForestClassifier(max_depth=depth, n_estimators=estimator)\n",
    "        scores = cross_val_score(model, X, y, cv=5)\n",
    "        model.fit(X_train, y_train)\n",
    "        y_pred = model.predict(X_val)\n",
    "        f1 = f1_score(y_true=y_val, y_pred=y_pred, average='weighted')\n",
    "        if f1 > best_model['f1']:\n",
    "            best_model = {'depth' : depth, 'n_estimators' : estimator, 'acc' : acc, 'f1' : f1, 'model' : copy.deepcopy(model)}\n",
    "            best_entry = pd.Series({'depth' : depth, 'n_estimator' : estimator, 'cv_score_fold1' : scores[0], 'cv_score_fold2' : scores[1], 'cv_score_fold3' : scores[2], 'cv_score_fold4' : scores[3], 'cv_score_fold5' : scores[4], 'mean' : scores.mean(), 'std' : scores.std(), 'accuracy' : acc, 'f1_score' : f1})\n",
    "    ALL_MODELS.append(best_entry)\n",
    "\n",
    "y_pred = best_model['model'].predict(X_val)\n",
    "print(f\"Best Model based on Mean Validation Accuracy : {best_model}\\nTest Accuracy : {accuracy_score(y_true=y_val, y_pred=y_pred)}\\nTest F1-Score : {f1_score(y_true=y_val, y_pred=y_pred, average='weighted')}\")\n",
    "\n",
    "name = f\"rdf_{best_model['depth']}_{best_model['n_estimator']}_{best_model['acc']:.2f}_{best_model['f1']:.2f}\"\n",
    "pickle.dump(best_model['model'], open(os.path.join(models_path, name), 'wb'))\n",
    "\n",
    "result = pd.DataFrame(ALL_MODELS)\n",
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>depth</th>\n",
       "      <th>n_estimator</th>\n",
       "      <th>cv_score_fold1</th>\n",
       "      <th>cv_score_fold2</th>\n",
       "      <th>cv_score_fold3</th>\n",
       "      <th>cv_score_fold4</th>\n",
       "      <th>cv_score_fold5</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>accuracy</th>\n",
       "      <th>f1_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.883101</td>\n",
       "      <td>0.882990</td>\n",
       "      <td>0.882990</td>\n",
       "      <td>0.883765</td>\n",
       "      <td>0.883170</td>\n",
       "      <td>0.000300</td>\n",
       "      <td>0.89572</td>\n",
       "      <td>0.833861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>400.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.883101</td>\n",
       "      <td>0.882990</td>\n",
       "      <td>0.884096</td>\n",
       "      <td>0.893055</td>\n",
       "      <td>0.885249</td>\n",
       "      <td>0.003925</td>\n",
       "      <td>0.89572</td>\n",
       "      <td>0.837013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>500.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.883322</td>\n",
       "      <td>0.884096</td>\n",
       "      <td>0.857222</td>\n",
       "      <td>0.878124</td>\n",
       "      <td>0.877154</td>\n",
       "      <td>0.010185</td>\n",
       "      <td>0.89572</td>\n",
       "      <td>0.861902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>300.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.883875</td>\n",
       "      <td>0.887746</td>\n",
       "      <td>0.759124</td>\n",
       "      <td>0.802588</td>\n",
       "      <td>0.843267</td>\n",
       "      <td>0.052804</td>\n",
       "      <td>0.89572</td>\n",
       "      <td>0.870560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>150.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.882216</td>\n",
       "      <td>0.872595</td>\n",
       "      <td>0.596771</td>\n",
       "      <td>0.730259</td>\n",
       "      <td>0.792969</td>\n",
       "      <td>0.113875</td>\n",
       "      <td>0.89572</td>\n",
       "      <td>0.871102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6.0</td>\n",
       "      <td>450.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.867728</td>\n",
       "      <td>0.869498</td>\n",
       "      <td>0.535833</td>\n",
       "      <td>0.608494</td>\n",
       "      <td>0.752911</td>\n",
       "      <td>0.149452</td>\n",
       "      <td>0.89572</td>\n",
       "      <td>0.874399</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7.0</td>\n",
       "      <td>350.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.849259</td>\n",
       "      <td>0.864521</td>\n",
       "      <td>0.522230</td>\n",
       "      <td>0.570228</td>\n",
       "      <td>0.737848</td>\n",
       "      <td>0.157554</td>\n",
       "      <td>0.89572</td>\n",
       "      <td>0.882252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8.0</td>\n",
       "      <td>300.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.839416</td>\n",
       "      <td>0.857885</td>\n",
       "      <td>0.515815</td>\n",
       "      <td>0.547113</td>\n",
       "      <td>0.728647</td>\n",
       "      <td>0.161895</td>\n",
       "      <td>0.89572</td>\n",
       "      <td>0.888941</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9.0</td>\n",
       "      <td>200.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.831232</td>\n",
       "      <td>0.850807</td>\n",
       "      <td>0.511060</td>\n",
       "      <td>0.454324</td>\n",
       "      <td>0.706085</td>\n",
       "      <td>0.184024</td>\n",
       "      <td>0.89572</td>\n",
       "      <td>0.894344</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10.0</td>\n",
       "      <td>500.0</td>\n",
       "      <td>0.883003</td>\n",
       "      <td>0.828467</td>\n",
       "      <td>0.836209</td>\n",
       "      <td>0.511944</td>\n",
       "      <td>0.336319</td>\n",
       "      <td>0.679189</td>\n",
       "      <td>0.216338</td>\n",
       "      <td>0.89572</td>\n",
       "      <td>0.897736</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   depth  n_estimator  cv_score_fold1  cv_score_fold2  cv_score_fold3  \\\n",
       "0    1.0        100.0        0.883003        0.883101        0.882990   \n",
       "1    2.0        400.0        0.883003        0.883101        0.882990   \n",
       "2    3.0        500.0        0.883003        0.883322        0.884096   \n",
       "3    4.0        300.0        0.883003        0.883875        0.887746   \n",
       "4    5.0        150.0        0.883003        0.882216        0.872595   \n",
       "5    6.0        450.0        0.883003        0.867728        0.869498   \n",
       "6    7.0        350.0        0.883003        0.849259        0.864521   \n",
       "7    8.0        300.0        0.883003        0.839416        0.857885   \n",
       "8    9.0        200.0        0.883003        0.831232        0.850807   \n",
       "9   10.0        500.0        0.883003        0.828467        0.836209   \n",
       "\n",
       "   cv_score_fold4  cv_score_fold5      mean       std  accuracy  f1_score  \n",
       "0        0.882990        0.883765  0.883170  0.000300   0.89572  0.833861  \n",
       "1        0.884096        0.893055  0.885249  0.003925   0.89572  0.837013  \n",
       "2        0.857222        0.878124  0.877154  0.010185   0.89572  0.861902  \n",
       "3        0.759124        0.802588  0.843267  0.052804   0.89572  0.870560  \n",
       "4        0.596771        0.730259  0.792969  0.113875   0.89572  0.871102  \n",
       "5        0.535833        0.608494  0.752911  0.149452   0.89572  0.874399  \n",
       "6        0.522230        0.570228  0.737848  0.157554   0.89572  0.882252  \n",
       "7        0.515815        0.547113  0.728647  0.161895   0.89572  0.888941  \n",
       "8        0.511060        0.454324  0.706085  0.184024   0.89572  0.894344  \n",
       "9        0.511944        0.336319  0.679189  0.216338   0.89572  0.897736  "
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "name = f\"rdf_{best_model['depth']}_{best_model['n_estimators']}_{best_model['acc']:.2f}_{best_model['f1']:.2f}\"\n",
    "pickle.dump(best_model['model'], open(os.path.join(models_path, name), 'wb'))\n",
    "\n",
    "result = pd.DataFrame(ALL_MODELS)\n",
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC Score : 0.676\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.roc_curve.RocCurveDisplay at 0x1ca49067a30>"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred = best_model['model'].predict(X_val)\n",
    "print(f\"AUC Score : {roc_auc_score(y_true=y_val, y_score=y_pred):.3f}\")\n",
    "RocCurveDisplay.from_predictions(y_true=y_val, y_pred=y_pred, name=\"ROC Curve\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Feature Importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>11</th>\n",
       "      <th>12</th>\n",
       "      <th>13</th>\n",
       "      <th>14</th>\n",
       "      <th>15</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.382703</td>\n",
       "      <td>-0.169813</td>\n",
       "      <td>-0.740666</td>\n",
       "      <td>2.012493</td>\n",
       "      <td>-0.13549</td>\n",
       "      <td>-0.235906</td>\n",
       "      <td>-0.893915</td>\n",
       "      <td>-0.436803</td>\n",
       "      <td>2.225971</td>\n",
       "      <td>-0.958642</td>\n",
       "      <td>-0.924841</td>\n",
       "      <td>0.636198</td>\n",
       "      <td>-0.246560</td>\n",
       "      <td>-0.411453</td>\n",
       "      <td>-0.251940</td>\n",
       "      <td>-0.415287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.559037</td>\n",
       "      <td>-0.169813</td>\n",
       "      <td>-0.740666</td>\n",
       "      <td>-0.051543</td>\n",
       "      <td>-0.13549</td>\n",
       "      <td>-0.299623</td>\n",
       "      <td>1.118674</td>\n",
       "      <td>-0.436803</td>\n",
       "      <td>0.406931</td>\n",
       "      <td>-0.239953</td>\n",
       "      <td>0.647631</td>\n",
       "      <td>-0.633582</td>\n",
       "      <td>-0.246560</td>\n",
       "      <td>-0.411453</td>\n",
       "      <td>-0.251940</td>\n",
       "      <td>-0.415287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.464863</td>\n",
       "      <td>0.152944</td>\n",
       "      <td>-0.740666</td>\n",
       "      <td>2.012493</td>\n",
       "      <td>-0.13549</td>\n",
       "      <td>-0.447419</td>\n",
       "      <td>1.118674</td>\n",
       "      <td>-0.436803</td>\n",
       "      <td>0.406931</td>\n",
       "      <td>-0.120171</td>\n",
       "      <td>0.018642</td>\n",
       "      <td>-0.078296</td>\n",
       "      <td>-0.246560</td>\n",
       "      <td>-0.411453</td>\n",
       "      <td>-0.251940</td>\n",
       "      <td>-0.415287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.795313</td>\n",
       "      <td>1.121214</td>\n",
       "      <td>-0.740666</td>\n",
       "      <td>-1.083560</td>\n",
       "      <td>-0.13549</td>\n",
       "      <td>0.950405</td>\n",
       "      <td>-0.893915</td>\n",
       "      <td>-0.436803</td>\n",
       "      <td>0.406931</td>\n",
       "      <td>-0.479516</td>\n",
       "      <td>0.018642</td>\n",
       "      <td>-0.074413</td>\n",
       "      <td>0.076230</td>\n",
       "      <td>7.198833</td>\n",
       "      <td>0.182198</td>\n",
       "      <td>2.448149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.665225</td>\n",
       "      <td>-0.815326</td>\n",
       "      <td>-0.740666</td>\n",
       "      <td>-0.051543</td>\n",
       "      <td>-0.13549</td>\n",
       "      <td>-0.447419</td>\n",
       "      <td>-0.893915</td>\n",
       "      <td>-0.436803</td>\n",
       "      <td>0.406931</td>\n",
       "      <td>-0.120171</td>\n",
       "      <td>-0.295852</td>\n",
       "      <td>-0.567569</td>\n",
       "      <td>-0.569351</td>\n",
       "      <td>-0.411453</td>\n",
       "      <td>-0.251940</td>\n",
       "      <td>-0.415287</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2         3        4         5         6  \\\n",
       "0  0.382703 -0.169813 -0.740666  2.012493 -0.13549 -0.235906 -0.893915   \n",
       "1 -0.559037 -0.169813 -0.740666 -0.051543 -0.13549 -0.299623  1.118674   \n",
       "2 -0.464863  0.152944 -0.740666  2.012493 -0.13549 -0.447419  1.118674   \n",
       "3  1.795313  1.121214 -0.740666 -1.083560 -0.13549  0.950405 -0.893915   \n",
       "4  0.665225 -0.815326 -0.740666 -0.051543 -0.13549 -0.447419 -0.893915   \n",
       "\n",
       "          7         8         9        10        11        12        13  \\\n",
       "0 -0.436803  2.225971 -0.958642 -0.924841  0.636198 -0.246560 -0.411453   \n",
       "1 -0.436803  0.406931 -0.239953  0.647631 -0.633582 -0.246560 -0.411453   \n",
       "2 -0.436803  0.406931 -0.120171  0.018642 -0.078296 -0.246560 -0.411453   \n",
       "3 -0.436803  0.406931 -0.479516  0.018642 -0.074413  0.076230  7.198833   \n",
       "4 -0.436803  0.406931 -0.120171 -0.295852 -0.567569 -0.569351 -0.411453   \n",
       "\n",
       "         14        15  \n",
       "0 -0.251940 -0.415287  \n",
       "1 -0.251940 -0.415287  \n",
       "2 -0.251940 -0.415287  \n",
       "3  0.182198  2.448149  \n",
       "4 -0.251940 -0.415287  "
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import os, pickle\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from matplotlib import pyplot as plt\n",
    "from collections import defaultdict\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.inspection import permutation_importance\n",
    "\n",
    "models_path = 'D:\\IIT JODHPUR\\Semester-3\\ML for Business\\Course Project\\\\best_models'\n",
    "all_models = os.listdir(models_path)\n",
    "\n",
    "df = pd.read_csv('Val_Standardized.csv')\n",
    "y = df['y']\n",
    "X = df.drop('y', axis=1)\n",
    "X = X.drop('Unnamed: 0', axis=1)\n",
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = np.array(X)\n",
    "y = np.array(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(16,)"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X[0].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_r = {}\n",
    "for m in all_models:\n",
    "    model = pickle.load(open(os.path.join(models_path, m), 'rb'))\n",
    "    r = permutation_importance(model, X, y, n_repeats=5)\n",
    "    model_r[m] = r"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = columns = ['age', 'job', 'marital', 'education', 'default', 'balance', 'housing',\n",
    "       'loan', 'contact', 'day', 'month', 'duration', 'campaign', 'pdays',\n",
    "       'previous', 'poutcome']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "algo = \"dtree_8_0.90_0.90\"\n",
    "importance = model_r[algo]\n",
    "\n",
    "plt.barh(range(len(importance['importances_mean'])), importance['importances_mean'], xerr=importance['importances_std'])\n",
    "plt.yticks(range(len(importance['importances_mean'])), columns)\n",
    "plt.ylabel('Features')\n",
    "plt.xlabel('Mean Importance +/- Mean SD')\n",
    "plt.title('Feature Improtance from Decision Tree')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "algo = \"rdf_10_500_0.90_0.90\"\n",
    "importance = model_r[algo]\n",
    "\n",
    "plt.barh(range(len(importance['importances_mean'])), importance['importances_mean'], xerr=importance['importances_std'])\n",
    "plt.yticks(range(len(importance['importances_mean'])), columns)\n",
    "plt.ylabel('Features')\n",
    "plt.xlabel('Mean Importance +/- Mean SD')\n",
    "plt.title('Feature Improtance from Random Forest')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAagAAAEWCAYAAAAzcgPFAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAAAxyUlEQVR4nO3deZxcVZn/8c+XBAgQSICETQzNJjsECCCbNsg4g7IvskNQQWQbxomQn6ATZFERB1EGEFAWWUSQTRAUkWYJRJJASAj7EmULhC1AwLDk+f1xTpOborq7utPVdTv9fb9e9cqtu5z7VFWnnjr3nvtcRQRmZmZls1CjAzAzM6vGCcrMzErJCcrMzErJCcrMzErJCcrMzErJCcrMzErJCcrMkLS1pKckvStpt0bH00rS9yRd1Og4rDGcoKxdkqZJej9/cbU+VuqGNnforhhr2N8YSZf31P7aI+kSSac2Oo4qfgicExEDI+KGntyxpCZJIal/5bKIOD0ivtmT8bRF0nqS/iLpDUlvSZoo6SuSPiPpI0mrV9nmekln5umQ9GrxdUpaOM/zBalVOEFZLXbOX1ytj5caGUy1L7KykNSv0TF00SrA1GoLlPSp74o2Psc/ArcDKwDLAccCb0fEi8AdwEEVbSwDfAW4tDD7TWDHwvMd8zyrok/90Vn3kTRI0q8lvSzpRUmntv6nlrS6pL9Jel3Sa5KukDQ4L/stMAz4Y+6NHS+pWdILFe1/0svKPaBrJV0u6W1gZHv7ryH2kHRkPqT1jqRTcsz3SXpb0u8lLZLXbZb0Qj7U9FqO64BCW5dIOk/SnyTNAraTtI6klvwre6qkXfK6hwMHAMfn1/7HPH+0pGdyLI9K2r3Q/khJ90o6U9Kbkp6TtGNh+TKSLpb0Ul5+Q2HZTpIm5Tjuk7RhG+/HM8Bqhc9k0Rz/aZLGAu8Bq0naStJ4STPzv1sV2mjJn8F9ra9N0rL5s387r99Uy+dTEdsnvV/N7WkdIumf+fM4sbDuQoX38vX8OS5TWH6NpOk5/rslrdfe51gRxxBgVeDCiPggP8ZGxL15lUupSFDAvsCjETGlMO+3wMGF5wcDl3X2fekzIsIPP9p8ANOAHarMvx74FbAE6dfkA8C38rI1gH8DFgWGAncDP2+rTaAZeKGt/QJjgA+B3Ug/qhZrb/9VYh0DXF54HsCNwFLAesBs0i/g1YBBwKPAIYXYPgL+N7+eLwKzgLXy8kuAmcDWObYlgaeB7wGLANsD71Ssf2pFfHsDK+Xt98ntr5iXjcyv/TCgH/Bt4CVAefktwNXA0sDCwBfz/I2BV4Et8naH5Pd00Vo+Z6AF+Gd+f/oDy5N+6R+Un++Xny9bWP9pYPXCe/gksENe/zLg4jb23ZQ/k/7tfXaF9S7MfwMb5c9unbz8P4FxwMr5s/oVcFWhra/nz2dR4OfApMKyys9xQEUcAp4Cbib9HS5fsXyxvP02hXn3A8dV/N2tD7wCDM6f2St5XjT6/3oZHw0PwI9yP/IX17vAW/lxQ/6ymg0sVlhvP+DONtrYDXioos3OJqi7C8s6u/9PvuTy8wC2LjyfCJxQeP4zckJlboJaorD898D38/QlwGWFZdsC04GFCvOuAsYU1j+1WpyF9ScBu+bpkcDThWWL5/hXAFYE5gBLV2njPOCUinlPkBNYG59zZYL6YeH5QcADFdvcD4wsrH9ixXt4a+H5zhQSQkU7TXQuQa1cWP4AsG+efgz4UmHZiqTkXq3dwbmtQdU+xzbiXBk4B3gmv+93A2sWll8EXJCn1wQ+AJar+LtbI6/3LeAIUrJdAyeoqo/SHsu3UtktIv7a+kTS5qRf6y9Lap29EPB8Xr48cDbpy3rJvGx+j7M/X5hepb391+iVwvT7VZ6vUHj+ZkTMKjz/B6nHUy22lYDnI2JOxfqfaSsQSQcD3yF9AQMMBIYUVpneOhER7+XXPBBYBngjIqq9t6sAh0g6pjBvkYq4O1L5uv5RsbzydXX0ng7sxL7bM70w/V6h3VWA6yUV3/uPgeUlTQdOI/VWh5ISDKT3eWaebvfvJyJeAI4GkPRZ4AJSz3DLvMqlwE2SjiUl9D9HxKtVmroM+BGpV3ZCu6+0j/M5KOuK50k9mCERMTg/loqI1mP6p5N+LW4QEUsBB5L+M7aqHLE0i9QzAD45QT20Yp3iNh3tv7stLWmJwvNhpMNs1WJ7Cfis5h1UMAx4scq6SFqF9Cv6aNLhssHAI8z7frXleWAZ5fN7VZadVnh/BkfE4hFxVQ3ttqp8XatULC++rjJ4Htix4jUPiDSIYX9gV9Ihx0HM/THQ3t9lmyLieeD/SIfnWt0LvJH3cyDzDo4ouofUu1s+b2NtcIKyTouIl4G/AD+TtFQ+Ob26pC/mVZYkHRacKekzwHcrmniFdL6n1ZPAAElflbQwcBLpPEFX918PJ0taRNK2wE7ANW2s93fSr/rjlYYQN5MOb/0uL6987UuQvhhnAEg6lHm/9NqU34dbgXMlLZ3394W8+ELgCElbKFkiv79L1vh6K/0J+Jyk/SX1l7QPsC7pnEx3WVTSgMKjs99P5wOn5aSPpKGSds3LliT9qHmd9GPo9M40nN/fkyWtkf/ehpDOaY1rXSfScbzLgJ+QDiH+sVpbeb2dgV3ytLXBCcq66mDSIaNHSYfvriX9KgQ4GdiEdOjkFuC6im1/BJyUR5eNioiZwJGkY/MvknpUL9C+9vbf3abnfbwEXAEcERGPV1sxIj4gffnsCLwGnAscXFj/18C6+bXfEBGPks7X3E9KXhsAYzsR20Gk8yyPkwZFHJfjmEAaWHFOjv1p0vmsLomI10mJ+b9JX/LHAztFxGtdbbOKd0mHAlsf23dy+7OBm4C/SHqHlDy2yMsuIx2SfJH0NzOuagtt+4DU6/or8DaplzubT7+nl5F6lldHxOy2GouIqRFRdVi/zSUncLO25R7Q5RGxcoNDMetz3IMyM7NScoIyM7NS8iE+MzMrJfegzMyslHyhbhcNGTIkmpqaGh2GmVmvMnHixNciovI6x6qcoLqoqamJCRMmNDoMM7NeRVJlRZI2+RCfmZmVkhOUmZmVkhOUmZmVkhOUmZmVkhOUmZmVkhOUmZmVkhOUmZmVkhOUmZmVki/U7UWaRt/S6BCsTqb9+KuNDsGsdNyDMjOzUip9gpLULGmrRsdhZmY9q/QJCmgGnKDMzPqYup2DktQE3AZMBDYBpgIHA1sCZ+Z9jwe+HRGzJU0DRkTEa5JG5HVGAkcAH0s6EDgGeBI4H1gt7+rbEXGfpO8AX8/zLoqInxdiGEdKcuOBi4GTgeWAAyLiAUlLAL8E1gcWBsZExI31eF/MAKZfOXqe583jfvqpdVpaWnooGrNyqncPai3g3IhYB3gb+A5wCbBPRGxASlLfbmvjiJhGSkZnRcTwiLgH+AVwV0RsRE58kjYFDgW2AD4PHCZp49zMGsDPgLXzY39gG2AU8L28zonA3yJic2A74Kc5ac1D0uGSJkiaMGPGjC6+JWZmVot6j+J7PiLG5unLge8Dz0XEk3nepcBRwM870eb2pJ4YEfExMFPSNsD1ETELQNJ1wLbATXl/U/L8qcAdERGSpgBNuc0vA7tIGpWfDwCGAY8VdxwRFwAXAIwYMcK3IrYuW2H/H8/zvMWj+Mw+pd4JqvJL/C1g2TbW/Yi5PboB3RjD7ML0nMLzOcx9/QL2jIgnunG/ZmY2H+p9iG+YpC3z9P7ABKBJ0hp53kHAXXl6GrBpnt6z0MY7wJKF53eQDwtK6idpEHAPsJukxfOhud3zvFr9GThGknK7G3ewvpmZ1Vm9E9QTwFGSHgOWBs4inSu6Jh9im0M6xwRp4MLZkiYAHxfa+COwu6RJkrYF/hPYLm8/EVg3Ih4kndt6APg7aZDEQ52I8xTS4IjJ+TDgKV16tWZm1m0UUZ9TKXkE3c0RsX5ddtBgI0aMCN/y3cyscyRNjIgRtazbG66DMjOzPqhugyTyEPEFsvdkZmb15x5UL+JisWbWlzhBmZlZKS1QCUrSnyQNbnQcZmY2/0p7PyhJ/XKliJpFxFfqFY+ZmfWshvSgJDVJelzSFZIek3Rtvsh2mqSfSHoQ2FvSlyXdL+lBSddIGijpPyRdU2irWdLNeXqapCF5+juSHsmP4wr7faSw7ShJY/L0sZIelTRZ0u968O2oSXNzM9OvHE1zc/MnDzOzBVkje1BrAd+IiLGSfgMcmee/HhGb5ERzHbBDRMySdAKp2OzpwAWSlsi19/YB5kkoFcVjBfxd0l3Am+3EMxpYNVdWH1xtBUmHA4cDDBs2rEsv2szMatPIBFVZSPbYPH11/vfzwLrA2FyBaBHg/oj4SNJtwM6SrgW+Chxf0XZ7xWPbMhm4QtINwA3VVmhksdiWlhaaRt/ioqJm1mc0MkFVfsG3Pp+V/xVwe0TsV2Xb3wFHA28AEyLinRr3WSxIC/MWpf0q8AVgZ+BESRtExEc1tmtmZt2skaP4KgvJ3luxfBywdWthWUlLSPpcXnYX6V5Qh1FxeC9rq3jsK8BykpaVtCiwU257IeCzEXEncAIwCBjYTa/TzMy6oJEJqrKQ7HnFhRExg3RH3askTQbuJ91wsPU+UDcDO+Z/qdi2avHYiPgQ+GGefzvweN6kH3B5LkD7EPCLiHirG1+rmZl1Ut2Kxba70wWgkKyLxZqZdZ6LxZqZWa/XkEESLiTbNY2sxTfNowfNrIe5B2VmZqW0QCWoYlUJMzPr3RaoBGVmZguO0haLrZRH/t0GTCRdAzUVOJh0ce3PgfcoXEslaXPgbNLFuO8Dh0bEE5LuBo6NiEl5vXuBo4DBeX1IFw1/oRMXAC+Qpl85+pPp5nE//WS6paWlAdGYWV/T23pQawHnRsQ6wNuk2nwXkqo/bAqsUFj3cWDbiNgY+AGphh/Ar0nXV5Ev/B0QEQ8Do4CjImI4qSzS+5U7l3S4pAmSJsyYMaP7X52ZmX2i1/Sgsmr1+56LiKcAJF1OLuZKqgZxqaQ1ST2ihfP8a4DvS/ou8HXSBb0AY4H/lXQFcF1EvFC580bW4muEFfb/8SfTrgFoZj2tt/WgKpPCoHbWPQW4M18MvDO57l5EvEeqIrEr8DXgijz/x8A3gcVIBWrX7t7QzcysM3pbgqqs3/dXoEnS6nlesbDsIODFPD2yop2LgF8A4yPiTQBJq0fElIj4CTCeXFbJzMwao7clqMr6fWeRDundkm9y+Gph3TOAH0l6iIpDmRExkXQO6+LC7OPyzQ0nAx8Ct9bvZZiZWUcaUouvK7qzfp+klYAWYO2ImNOVNlyLz8ys81yLrx2SDiZVOD+xq8nJzMzqr9eM4uuu+n0RcRlw2XwHZGZmddVrElRfVSwQ64KtZtaX9LlDfGZm1jss0AlK0i6SRne8ppmZlc0CfYgvIm4Cbmp0HGZm1nl1TVB5xNwoUgWIycDvgZOARYDXgQMi4hVJY4BVgdWAYcB/AZ8HdiRdbLtzRHwoaVpuY0dSrbz9I+JpSTu30e5IYEREHJ0v5r0CWAK4ETguIgZKagbGAK+RBmFMBA6MBo2/b25unuf59Gdfn7ts3E9dqNXM+oy6HeKTtB4paWwfERsB/0mqNv75XMD1d8DxhU1WB7YHdiHV2bszIjYgJaLi6ICZef45pCrmdNBuq7OBs/O2lXX2NgaOA9YlJcmt23hNLhZrZtZD6tmD2h64JiJeA4iINyRtAFwtaUVSb+e5wvq35l7SFKAf6dYaAFOApsJ6VxX+PStPr9xOu622BHbL01cCZxaWPdBaHFbSpLy/e6nQE8ViK3tIxVF8LthqZn1JTw+S+CVwTu7FfItcwDWbDZAvnv2wcIhtDvMm0qgy3V67tZhdmP6YBfzcnJlZb1DPBPU3YG9JywJIWoZ5C7ge0sV29yn8e3+erqXdccCeeXrfLu7bzMx6SN16ChExVdJpwF2SPgYeIg1GuEbSm6QEtmoXml46F3Sdzdzq5bW0exxwuaQTSYcPZ3Zh32Zm1kN6TbFYgDyKb0Trea1Obrs48H5EhKR9gf0iYteuxuJisWZmndeZYrF96VzLpsA5kgS8RbqbrpmZlVSvSlAR0TQf294DbNR90fS84og+M7N6KkPtzwW61JGZmfVepUhQksZIGtUN7QyWdGTh+UqSrp3fds3MrOeVIkF1hqT2DksOBj5JUBHxUkTsVfegzMys2zXsHFQe7n0I8CrwPDBRUgswKiImSBoCTIiIplxTbw9gINBP0ldJ9fSWBhYGToqIG4EfA6vnahC3A/9Hvk28pAHAecAI4CPgOxFxZ257F2BxUrml6yOiWqkkM7MFzvQrq9/woXncT9vcpqdqgjYkQUnalHSx7PAcw4OkIq3t2QTYMJdM6g/sHhFv50Q2TtJNwGhg/YgYnvfTVNj+KCAiYgNJawN/kfS5vGw4qR7fbOAJSb+MiOerxH04cDjAsGHDOv26zcysdo3qQW1L6qm8B5CTS0duj4g38rSA0yV9gVQK6TPA8h1svw2pJBIR8bikfwCtCeqOiJiZY3kUWIXUq5tHT9TiMzPrSSvs/+Oq88tQ+7Nsw8w/Yu55scp6erMK0wcAQ4FNC7fh6Gz9vSLX4jMzK5lGDZK4G9hN0mKSlgR2zvOnkS6oBWhvcMMg4NWcnLYj9XgA3gGWbGObe0iJjXxobxjwRJdfgZmZ1VVDElREPAhcDTwM3AqMz4vOBL4t6SFgSDtNXAGMyLfmOBh4PLf7OjBW0iOSKs/wnQsslLe5GhgZEbMxM7NS6lW1+MrEtfjMzDqvM7X4et11UGZm1jc4QZmZWSl5tFoJtVUUtgzFG83Meop7UGZmVkp9MkFVKSrbLOnmRsZkZmbz6pMJioqismZmVj6lPweV6+ndBowDtiJdM3UxcDKwHOni26eB3wCrAe8Bh0fEZEljSBfkrpb//XlE/IJPF5W9BRiYb82xPqku4IHRA2Pwm5ubPzVv+rOvf2peW+VIzMwWVKVPUNkawN6k27SPB/Yn1dbbBfgeqW7eQxGxm6TtgctIBWAB1ga2I1WYeELSeXy6qGwzqVjsesBLwFhga+DeYhAuFmtm1nN6S4J6LiKmAEiaSiruGrkqRBOp1NGeABHxN0nLSloqb3tLrhgxW9KrtF1U9oGIeCHvY1Jud54EVY9isdXK1vvW7mZmveccVLEk0ZzC8zl0nGRrLQTrgrFmZiXSWxJUR4qFYJuB1yLi7XbWb6+orJmZlcCC0ksYA/xG0mTSIIlD2ls5Il6XNFbSI6RitT6mZmZWMi4W20UuFmtm1nkuFmtmZr3egnKIb4FVHNHnWnxm1pe4B2VmZqXUJxKUpDGSRjU6DjMzq12fSFBmZtb7LLAJStKJkp6UdC+wVp53mKTxkh6W9AdJi0taUtJzkhbO6yxVfN4ozc3NNDc3M/3K0Y0Mw8ysYWpKUJJWl7Ronm6WdKykwXWNbD5I2hTYl1SP7yvAZnnRdRGxWURsBDwGfCMi3gFagNYRCPvm9T6s0u7hkiZImjBjxow6vwozs76t1h7UH4CPJa1BqkX3WeDKukU1/7YFro+I93JFiZvy/PUl3ZNr+B1AKg4LcBFwaJ4+lFQt/VMi4oKIGBERI4YOHVrH8FONvpaWFlcxN7M+q9YENSciPgJ2B34ZEd8FVqxfWHVzCXB0RGxAul3HAICIGAs05TJJ/SLikUYFaGZmSa0J6kNJ+5FKCLXeebah52g6cDewm6TFJC0J7JznLwm8nM8vHVCxzWWkXmHV3pOZmfWsWhPUocCWwGkR8ZykVYHf1i+s+RMRDwJXAw+Tau2Nz4u+D/yddL+nxys2uwJYGriqh8I0M7N21FyLT9JiwLCIeKK+ITWGpL2AXSPioFrWdy0+M7PO6/ZafJJ2BiaRbr2OpOGSbmp3o15E0i9Jt4E/pdGxmJlZUmstvjHA5qTh2ETEJEmr1SmmHhcRxzQ6BjMzm1etCerDiJgpqThvTh3iMdq+5buLxZpZX1JrgpoqaX+gn6Q1gWOB++oXlpmZ9XW1juI7hnRR62zSUOyZwHF1iqnLJDXlRNrV7UdKWqk7YzIzs67psAclqR9wS0RsB5xY/5DmSxOwP12vcjESeAR4qZviMTOzLuowQUXEx5LmSBoUETPrGYykg4FRQACTSdct/QYYAswADo2If0q6BHgbGAGsABwfEdeSRuKtI2kScClwPel6rSXyLo6OiPvyvk4ADiSdS7sVmJDbu0LS+8CWEfF+PV9vc3Nz1fnTn339U/Nc8sjM+ppaz0G9C0yRdDswq3VmRBzbXYFIWg84CdgqIl6TtAwpyVwaEZdK+jrwC2C3vMmKwDbA2qRae9cCo4FREbFTbnNx4N8i4l/53NlVwAhJOwK7AltExHuSlomINyQdnbeveoGTpMOBwwGGDRvWXS/dzMyqqDVBXZcf9bQ9cE1EvAaQE8aWwB55+W+BMwrr3xARc4BHJS3fRpsLA+dIGg58DHwuz98BuDgi3mvdVy0BRsQFpGK5jBgxorYrnNvR0tJSdX5bo/jMzPqSmhJURFxa70C6YHZhWm2s81/AK8BGpAEh/6p3UGZm1j1qrSTxnKRnKx/dHMvfgL0lLZv3uQxpKPu+efkBwD0dtPEOqSBsq0HAy7mndRDQL8+/HTg0HwJs3Ve17c3MrEFqPcRXrJs0ANgbWKaNdbskIqZKOg24S9LHwEOk4e0XS/oueZBEB81MJt236mHSrTXOBf6QB1/cRj5/FhG35cN+EyR9APwJ+F7e5vyeGiRhZmZtq7lY7Kc2TAX/Nu3meHoNF4s1M+u8zhSLrakHJWmTwtOFSD2qWntfZmZmnVZrkvlZYfoj4Dnga90fjrWnafQtrsdnZn1GrQnqGxExz6CIfNNCMzOzuqi1Ft+1Nc5rKEnvNjoGMzPrHu32oCStTSoSO0jSHoVFS5FG85mZmdVFR4f41gJ2AgYDOxfmvwMcVqeY5pvSjavOAHYk1fU7NSKuljQQuBFYmlRl4qSIuFFSE6ke373AVsCLpNu/N3yYebFe3/RnX6d53E8/ed5WJQozswVBuwkqIm4EbpS0ZUTc30MxdYc9gOGkChJDgPGS7iZdS7V7RLwtaQgwrnDr+jWB/SLiMEm/B/YELi826lp8ZmY9p9ZBEg9JOop0uO+TQ3sR8fW6RDX/tgGuioiPgVck3QVsRuolnS7pC6Qq5p8BWuv4PRcRk/L0RNKtO+bR3bX4alHsJTWNvoUWj+Izsz6i1kESvyXd1uLfgbuAlUmH+XqbA4ChwKYRMZxUp6814RZr+32Mr/MyM2uoWhPUGhHxfWBWLhz7VWCL+oU13+4B9pHUT9JQ4AvAA6TafK9GxIeStgNWaWSQZmbWtlp7CR/mf9+StD4wHViuPiF1i+uBLYGHSYMkjo+I6ZKuAP4oaQrpBoWPNzBGMzNrR021+CR9E/gDsCFwMTAQ+EFEnF/f8MrLtfjMzDqv22vxRcRFefIuYLWuBmZmZlarWu8HtbykX0u6NT9fV9I36huamZn1ZbWeg7qEdGjvxPz8SeBq4Nd1iMna0JO3gndRWjNrtFpH8Q2JiN+Trh0iIj4iDcU2MzOri1oT1Kx8K/YAkPR5YGZ3BCCpSdIj3dFWG+3fV6+2zcysfmo9xPcd4CZgdUljSRe77lW3qLpRRGzV6BjMzKzzOqpmPiwi/hkRD0r6Iql4rIAnIuLD9rbtpH6SLqRQqDXv63xgceAZ4OsR8aakFmBUREzI9fQmRESTpPVI58kWIfUM94yIpyS9GxEDJTUDY4DXgPVJ5YwOjIiQ9BXgf4FZwFhgtYjYqRtfX6lNv3L0p+YVi9KCC9OaWc/r6BDfDYXpqyNiakQ80s3JCVKh1v+LiPWAt0iFWi8DToiIDYEpwP900MYRwNm5hNEI4IUq62wMHAesSxouv7WkAcCvgB0jYlNS77AqSYdLmiBpwowZM2p/dWZm1mkdHeJTYbqe1z9VFmpdHRgcEXfleZcC13TQxv3AiZJWBq6LiKeqrPNARLwAIGkSqSDsu8CzEfFcXucqcsXySo0oFtsTVtj/x5+a56K0ZtZoHfWgoo3p7lZZqHVwO+t+xNy4i5XVrwR2Ad4H/iRp+xr244KwZmYl1VGC2kjS25LeATbM029LekfS23WMaybwpqRt8/ODSFUsAKYBm+bpTwZqSFqN1BP6BemmhBvWuK8ngNXyTQsB9ul62GZm1l06umFhv54KpIpDgPMlLQ48Cxya558J/D7fPLB45erXgIMkfUgqZnt6LTuJiPclHQncJmkWML67XoCZmXVdTcViF3SSBkbEu/lW8f8HPBURZ7W3jYvFmpl1XmeKxdZ6oe6C7rA8aGIq6Z5Rv2psOGZm5kECQO4ttdtj6klt1dxzfTwz60vcgzIzs1JqeILqbC0+SZdI6hVllszMrOsanqDMzMyqKUuC6i/pCkmPSbpW0uKSfiBpvKRHJF2QR9jNo611JLVI+omkByQ92Xo9laR+ks7M60+WdEyev6mkuyRNlPRnSSv27Mufq7m5uWptPDOzvqYsCWot4NyIWAd4GzgSOCciNouI9YHFgGrFW9tbp39EbE6qvddax+9wUnmj4bnG3xWSFgZ+CeyVa/H9BjitWpCuxWdm1nPKkqCej4ixefpyYBtgO0l/lzQF2B5Yr8p27a1zXf53IikpAewA/CrfcJGIeIOUHNcHbs9DzU8CVq4WZERcEBEjImLE0KFt1pSdLy0tLVVr45mZ9TVlGWZeebVwAOcCIyLieUljKNTdA8hVyNtbp7XuXkc19wRMjYgtux6+mZl1t7L0oIZJak0Q+wP35unXJA2k+s0RB9SwTqXbgW9J6g8gaRlSLb6hrfuXtHC+t5SZmTVQWRLUE8BRkh4DlgbOAy4EHgH+TJX6eBHxVkfrVHER8E9gsqSHgf0j4gNScvtJnjeJdONEMzNrINfi6yLX4jMz6zzX4jMzs17PCcrMzEqpLKP4rA3FwrEuFmtmfYl7UGZmVkqlT1CSxkga1c7yofli3YcKt4jvTPsjJZ2Tp3eTtO78xGtmZt2j9AmqBl8CpkTExhFxz3y2tRvgBGVmVgKlPAcl6UTgEOBV4HlgoqTVSbdjHwq8BxxGulj3DGAxSSOALYH/BTYj1ea7NiL+J7c5jVR14rW87pkR0VzY51bALsAXJZ0E7BkRz/TAy/2U5uZPwmL6s68DuPyRmfU5pUtQkjYF9gWGk+J7kFRP7wLgiIh4StIWpOKy20v6ASnxHJ23PzEi3pDUD7hD0oYRMbmj/UbEfZJuAm6OiGvbiO1wUsFZhg0bNt+v1czM2la6BAVsC1wfEe8B5KQxgFTd4ZrCXTcWbWP7r+VE0h9YkXTIrsMEVYuIuICUKBkxYkTdrnBuaWn5ZLqt27+bmS3oypigqlkIeCsihre3kqRVgVHAZhHxpqRLmFuz7yPmnnMbUGVzMzMrkTIOkrgb2E3SYpKWBHYmnXN6TtLeAEo2qrLtUsAsYKak5YEdC8umAZvm6T3b2Pc7wJLz/xLMzGx+lS5BRcSDwNXAw8CtzC0CewDwjVzQdSqwa5VtHwYeAh4HrgTGFhafDJwtaQLpFhzV/A74bh6yvno3vBwzM+siF4vtIheLNTPrPBeLNTOzXq+3DJLoU9oauedafGbWl7gHZWZmpdTQBFWsg9eNbc5TT0/SDyXt0J37MDOz+lsQe1C7UainFxE/iIi/Ni4cMzPrirqeg5J0IHAssAjwd+BI4GDg/wFvkYaSz87rXkKhzJCkdyNiYJ4+ATgQmAPcGhGjJR1GKju0CPA0cBCpPNI89fSA77e2K+lLwJn5dY8Hvh0Rs3OdvktJ11wtDOwdEY/X7Y0pKNbda9Vaf+9TfA7KzPqQuvWgJK0D7ANsnStAfExKMicDWwPbUEPlcEk7kq552iIiNiIVhwW4LiI2y/MeA74REfcBNwHfjYjhxWKvkgYAlwD7RMQGpCT17cKuXouITYDzSNUoqsVyuKQJkibMmDGjxnfCzMy6op49qC+RKjeMz/XzFiPV02uJiBkAkq4GPtdBOzsAF7fW5ouIN/L89SWdCgwGBgJ/7qCdtYDnIuLJ/PxS4Cjg5/n5dfnficAe1RqoRy2+Yt29Vq6/Z2ZW33NQAi7NPZnhEbEWMKad9T+plSdpIdKhu/ZcAhyde0MnM//19Wbnfz/Gw+/NzBqungnqDmAvScsBSFqGVIboi5KWlbQwsHdh/WnMrZW3C+lcEMDtwKGSFi+0A6lm3su5nQMK7bRVT+8JoEnSGvn5QcBdXX95ZmZWT3VLUBHxKHAS8BdJk0mJZkVSL+p+Up28xwqbXEhKXg+Tbjw4K7dzG+m80gRJk5h7fuj7pIEXY0m191pVracXEf8CDiXdsmMKacDF+d34ks3MrBu5Fl8XuRafmVnnuRafmZn1ek5QZmZWSh6tVjLtDTF3sVgz60vcgzIzs1JaoBOUpJUktZZOGi7pKzVs0yzp5vpHZ2Zm7VlgE5Sk/hHxUkTslWcNBzpMUGZmVg6lOwclqQm4DRhHKo00HriYVC1iOeZelHs2qXrE+8ChEfGEpJGkMkUDgX6SDgFuBjYBfggsJmkb4EfAc9Xa6IGXWLVAbKs2C8WCi8WaWZ9SugSVrUGqMvF1UoLan1Rcdhfge6SK6NtGxEf5Xk+nkyqXQ0pGG0bEGznZEREfSPoBMCIijgaQtFQ7bVQl6XBSBXWGDRvWjS/XzMwqlTVBPRcRUwAkTQXuiIjIFSCagEHApZLWBIK5ZZEAbi8UlG1Pe21U1V3FYqsViG3lQrFmZklZz0HNLkzPKTyfQ0qqpwB3RsT6pHs4FQvFzqpxH+21YWZmDVbWBNWRQcCLeXpkjdtUFpHtShtmZtZDemuCOgP4kaSHqP0w5Z3AupImSdqni22YmVkPcbHYLnKxWDOzznOxWDMz6/WcoHoRj/Azs77ECcrMzEqpzyQoSfe1s8z198zMSqbPJKiI2KrRMZiZWe36zPBqSe+SroM6A9iRVD3i1Ii4Oq+ylKRbSGWW7gSOjIg5DQm2iubmZqY/+zrN4376ybz2KlKYmfV2fSZBZXuQqppvBAwBxku6Oy/bHFgX+AepWO0ewLXFjV2Lz8ys5/S1BLUNcFVEfAy8IukuYDPgbeCBiHgWQNJVed15ElR31eLripaWFppG30KLK5qbWR/RZ85B1aAy4fgKZjOzBuprCeoeYB9J/SQNBb4APJCXbS5pVUkLAfsA9zYqSDMz61sJKoDrgcnAw8DfgOMjYnpePh44B3iMdDPD6xsRpJmZJX2iFp+kZYEHI2KV7mrTtfjMzDrPtfgKJK0E3A+c2ehYzMysdgv8KL6IeAn4XKPjMDOzzlngE1RvVlkcdpqHmJtZH7LAH+IzM7PeyQnKzMxKyQnKzMxKaYFNUJJukDRR0tRcQw9J35D0pKQHJF0o6Zw8f6ikP0ganx9bNzb6XBz2ytGNDsPMrGEW5EESX4+INyQtRioKewvwfWAT4B3ShboP53XPBs6KiHslDQP+DKxT2aCLxZqZ9ZwFOUEdK2n3PP1Z4CDgroh4A0DSNcwdfr4DsK6k1m2XkjQwIt4tNtiTxWJbi8OamfVVC2SCktRMSjpbRsR7klqAx6nSK8oWAj4fEf/qkQDNzKxDC+o5qEHAmzk5rQ18HlgC+KKkpSX1B/YsrP8X4JjWJ5KG92SwZmb2aQtqgroN6C/pMeDHwDjgReB0UvXyscA0YGZe/1hghKTJkh4FjujxiM3MbB59olhsq9bzSrkHdT3wm4joUtVyF4s1M+s8F4tt2xhJk4BHSLfUuKGh0ZiZWZsWyEESbYmIUY2OwczMatPXelBmZtZLOEGZmVkpOUGZmVkpOUGZmVkpOUGZmVkpOUGZmVkpOUGZmVkp9alKEt1J0gzgHz282yHAaz28z65yrN2vt8QJjrVeekus7cW5SkQMraURJ6heRNKEWkuENJpj7X69JU5wrPXSW2Ltrjh9iM/MzErJCcrMzErJCap3uaDRAXSCY+1+vSVOcKz10lti7ZY4fQ7KzMxKyT0oMzMrJScoMzMrJSeokpD0H5KekPS0pNFVli8q6eq8/O+SmgrL/l+e/4Skfy9rrJKWlXSnpHclnVPiOP9N0kRJU/K/25c41s0lTcqPhyXtXtZYC8uH5b+But6fbT7e0yZJ7xfe1/PrGef8xJqXbSjpfklT89/sgDLGKumAwns6SdIcScPb3VlE+NHgB9APeAZYDVgEeBhYt2KdI4Hz8/S+wNV5et28/qLAqrmdfiWNdQlgG+AI4JwSv6cbAyvl6fWBF0sc6+JA/zy9IvBq6/OyxVpYfi1wDTCqjHECTcAj9fzMuzHW/sBkYKP8fNmy/v+vWGcD4JmO9uceVDlsDjwdEc9GxAfA74BdK9bZFbg0T18LfEmS8vzfRcTsiHgOeDq3V7pYI2JWRNwL/KuO8XVHnA9FxEt5/lRgMUmLljTW9yLiozx/AFDvUU/z87eKpN2A50jva2nj7GHzE+uXgckR8TBARLweER+XNNai/fK27XKCKofPAM8Xnr+Q51VdJ38hzST9Wqpl2+40P7H2pO6Kc0/gwYiYXac454kj61SskraQNBWYAhxRSFililXSQOAE4OQ6xjffceZlq0p6SNJdkrYtcayfA0LSnyU9KOn4EsdatA9wVUc769/lMM0WcJLWA35C+pVaWhHxd2A9SesAl0q6NSJ6opfaWWOAsyLi3cZ0VGr2MjAsIl6XtClwg6T1IuLtRgdWRX/SYfPNgPeAOyRNjIg7GhtW2yRtAbwXEY90tK57UOXwIvDZwvOV87yq60jqDwwCXq9x2+40P7H2pPmKU9LKwPXAwRHxTJljbRURjwHvks6b1cv8xLoFcIakacBxwPckHV22OPPh8tcBImIi6ZzL5+oU53zFSurB3B0Rr0XEe8CfgE1KGmurfamh9wR4kEQZHqRfQc+SBjm0nnhcr2Kdo5j3xOPv8/R6zDtI4lnqe5K0y7EWlo+k/oMk5uc9HZzX36MXfP6rMneQxCrAS8CQMsZasc4Y6jtIYn7e06Gt/4dIgwFeBJYpaaxLAw+SB8sAfwW+WsZY8/OF8vu5Wk37q9cL8aPTH/xXgCdJv9ZOzPN+COySpweQRj49DTxQ/ICBE/N2TwA7ljzWacAbpF/6L1AxAqgMcQInAbOASYXHcmV8T4GDSAMOJuUvqt3K/PkX2hhDHRPUfL6ne1a8pzuX+T0FDszxPgKcUfJYm4Fxte7LpY7MzKyUfA7KzMxKyQnKzMxKyQnKzMxKyQnKzMxKyQnKzMxKyQnKegVJIenywvP+kmZIurnO+71E0l713EeVfR4nafGe3GdnKVV5b7M+Ya4IHpJOLcwbIulD1bmSvaSdcpmihyU9Kulbef4YSS/mStpPSbpO0rr1jMXmjxOU9RazgPUlLZaf/xv1rZjREJL6kaosNDxB5YoP1eavSqrw3lF9wueArxae702di8RKWph0u/GdI2IjUmX6lsIqZ0XE8IhYE7ga+JukofWMybrOCcp6kz8x9wtvPwrlUiQtIek3kh7Iv553zfObJN2TC2k+KGmrPL9ZUoukayU9LumKjipZS5om6Uf5F/gESZvkIp3PSDqi0O7dkm7J98w5X9JCedl+SvfreUTSTwrtvivpZ5IeJl10vRJwp6Q78/Lz8v6mSjq5Ip6T8+uaImntPH+gpIvzvMmS9szzv6x036AHJV2Ti7d2xX8At9Ww3nvAY5JG5Of7AL8vxD9U0h8kjc+PrfP8zXOcD0m6T9Jaef7I3Ou5LfeAzqiyzyVJ1Q5aSxXNjognqgUXEVcDfwH2r+1lW4+r91XHfvjRHQ9S5YkNSeX7B5Cu8m8Gbs7LTwcOzNODSVe6L0HqiQzI89cEJuTpZlKV5ZVJP9TuB7apst9LgL3y9DTg23n6LNJ9eJYklcZ5pdDuv0glcvoBtwN7kZLOP/O6/YG/kas+kG6R8bXCPqdRKFdELrOT22sBNiysd0yePhK4KE//BPh5YfulgSHA3cASed4JwA86eM+ntTH/RjooVUO+pxKwC3AmqTbbHRTKXAFXtr7nwDDgsTy9FHPLN+0A/CFPjySV2RmU/wb+AXy2yr4vIt0X6yrgAGChPH8MFdUrSL3V8xr99+1H9YermVuvERGTle7OuR+pN1X0ZWAXzb1L6wDSl95LwDlKd+78mHmLfj4QES8ASJpE+lK9t4Mwbsr/TgEGRsQ7wDuSZksaXGj32dzuVaRq0x8CLRExI8+/AvgCcEOO6w/t7PNrkg4nJbYVSTepnJyXXZf/nQjskad3INVAAyAi3pS0U95ubO4oLkJKyvOQdCLpUBzASvl9ARgbEUdJWgRYufX11eA24BTgFdIhtaIdgHULHdelcq9uEKkq+5qk5L1wYZs7ImJmjvVRUv3B4u0fiIhvStogtz+KdDh4ZBvxlbqsel/nBGW9zU2kX+TNzHuPGQF7RsXhHEljSF+OG5F6SsXbUBTPoXxMbf8fWreZU7H9nML2lfXDOqon9q9o4yZz+XzPKGCznGguISXfyng6il/A7RGxX3uBRMRpwGl539MiYnjFKtuSk7jS7eX/J8//ZkRMqNLeB5ImAv9NSpC7FBYvBHw+Km4NkgdR3BkRu+cfJC2FxTV9ZhExBZgi6bekc2Ejq79iNgY+FbeVg89BWW/zG+Dk/AVU9GfgmNbzSJI2zvMHAS9HxBxSYdV+PRDj5pJWzeee9iF9oT8AfFFpJFs/Ui/wrja2f4d06BDS4a5ZwExJywM71rD/20kVpQGQtDQwDtha0hp53hKSunILif8AbgWIiOsjDTgYXi05FfwMOCEi3qiY/xfgmEKcw/PkIOYOgBnZmeDy+bfmwqzhpEOB1dbdk9Tzru3WD9bjnKCsV4mIFyLiF1UWnUI6FDRZ6e6yp+T55wKH5AEIa5O+7OttPHAO8Bjp1/v1EfEyMBq4k3SLgokRcWMb218A3Cbpzki38n4IeJx0zmZsDfs/FVg6D8Z4GNguH1ocCVwlaTLp8N7aXXhtzbSdWKuKiKkRcWmVRccCI/JAjkeBI/L8M4AfSXqIzh/lEXB8HqAyiXT33pGF5f+VB7k8RaoCvn3rYVcrH1czN+tG+df7qIjYqcGhdDulmzheGBG19OLM5pvPQZlZTfKAEicn6zHuQZmZWSn5HJSZmZWSE5SZmZWSE5SZmZWSE5SZmZWSE5SZmZXS/wcLuZJ+vtK5MAAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "algo = \"svm_linear_10_0.86_0.87\"\n",
    "importance = model_r[algo]\n",
    "\n",
    "plt.barh(range(len(importance['importances_mean'])), importance['importances_mean'], xerr=importance['importances_std'])\n",
    "plt.yticks(range(len(importance['importances_mean'])), columns)\n",
    "plt.ylabel('Features')\n",
    "plt.xlabel('Mean Importance +/- Mean SD')\n",
    "plt.title('Feature Improtance from Linear SVM')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "algo = \"svm_poly_0.001_0.90_0.89\"\n",
    "importance = model_r[algo]\n",
    "\n",
    "plt.barh(range(len(importance['importances_mean'])), importance['importances_mean'], xerr=importance['importances_std'])\n",
    "plt.yticks(range(len(importance['importances_mean'])), columns)\n",
    "plt.ylabel('Features')\n",
    "plt.xlabel('Mean Importance +/- Mean SD')\n",
    "plt.title('Feature Improtance from SVM Polynomial Kernel')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "algo = \"svm_rbf_10_0.84_0.86\"\n",
    "importance = model_r[algo]\n",
    "\n",
    "plt.barh(range(len(importance['importances_mean'])), importance['importances_mean'], xerr=importance['importances_std'])\n",
    "plt.yticks(range(len(importance['importances_mean'])), columns)\n",
    "plt.ylabel('Features')\n",
    "plt.xlabel('Mean Importance +/- Mean SD')\n",
    "plt.title('Feature Improtance from SVM RBF Kernel')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "algo = \"logreg_0.001_0.83_0.85\"\n",
    "importance = model_r[algo]\n",
    "\n",
    "plt.barh(range(len(importance['importances_mean'])), importance['importances_mean'], xerr=importance['importances_std'])\n",
    "plt.yticks(range(len(importance['importances_mean'])), columns)\n",
    "plt.ylabel('Features')\n",
    "plt.xlabel('Mean Importance +/- Mean SD')\n",
    "plt.title('Feature Improtance from Logistic Regression')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>svm_poly</th>\n",
       "      <th>svm_rbf</th>\n",
       "      <th>linsvm</th>\n",
       "      <th>Dtree</th>\n",
       "      <th>RDF</th>\n",
       "      <th>LogReg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.0013 +/- 0.0005</td>\n",
       "      <td>0.0014 +/- 0.0014</td>\n",
       "      <td>-0.0010 +/- 0.0002</td>\n",
       "      <td>0.0048 +/- 0.0004</td>\n",
       "      <td>0.0018 +/- 0.0007</td>\n",
       "      <td>0.0000 +/- 0.0002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0000 +/- 0.0005</td>\n",
       "      <td>0.0030 +/- 0.0018</td>\n",
       "      <td>0.0003 +/- 0.0005</td>\n",
       "      <td>0.0000 +/- 0.0002</td>\n",
       "      <td>0.0008 +/- 0.0005</td>\n",
       "      <td>0.0017 +/- 0.0011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0003 +/- 0.0005</td>\n",
       "      <td>-0.0011 +/- 0.0006</td>\n",
       "      <td>-0.0016 +/- 0.0017</td>\n",
       "      <td>-0.0005 +/- 0.0004</td>\n",
       "      <td>0.0020 +/- 0.0005</td>\n",
       "      <td>-0.0001 +/- 0.0009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.0000 +/- 0.0003</td>\n",
       "      <td>0.0002 +/- 0.0010</td>\n",
       "      <td>-0.0014 +/- 0.0013</td>\n",
       "      <td>-0.0003 +/- 0.0001</td>\n",
       "      <td>0.0010 +/- 0.0003</td>\n",
       "      <td>0.0007 +/- 0.0015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0000 +/- 0.0004</td>\n",
       "      <td>0.0001 +/- 0.0006</td>\n",
       "      <td>-0.0008 +/- 0.0005</td>\n",
       "      <td>0.0000 +/- 0.0000</td>\n",
       "      <td>0.0000 +/- 0.0001</td>\n",
       "      <td>0.0001 +/- 0.0003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.0001 +/- 0.0003</td>\n",
       "      <td>0.0013 +/- 0.0012</td>\n",
       "      <td>-0.0015 +/- 0.0003</td>\n",
       "      <td>-0.0006 +/- 0.0009</td>\n",
       "      <td>0.0005 +/- 0.0007</td>\n",
       "      <td>-0.0010 +/- 0.0004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.0031 +/- 0.0005</td>\n",
       "      <td>0.0106 +/- 0.0012</td>\n",
       "      <td>0.0062 +/- 0.0011</td>\n",
       "      <td>0.0061 +/- 0.0007</td>\n",
       "      <td>0.0022 +/- 0.0006</td>\n",
       "      <td>0.0080 +/- 0.0010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.0004 +/- 0.0003</td>\n",
       "      <td>0.0012 +/- 0.0006</td>\n",
       "      <td>0.0001 +/- 0.0008</td>\n",
       "      <td>0.0002 +/- 0.0001</td>\n",
       "      <td>0.0005 +/- 0.0006</td>\n",
       "      <td>0.0006 +/- 0.0011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-0.0013 +/- 0.0006</td>\n",
       "      <td>0.0468 +/- 0.0036</td>\n",
       "      <td>-0.0015 +/- 0.0013</td>\n",
       "      <td>0.0178 +/- 0.0008</td>\n",
       "      <td>0.0044 +/- 0.0006</td>\n",
       "      <td>-0.0049 +/- 0.0013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.0001 +/- 0.0003</td>\n",
       "      <td>0.0037 +/- 0.0010</td>\n",
       "      <td>-0.0008 +/- 0.0003</td>\n",
       "      <td>0.0004 +/- 0.0005</td>\n",
       "      <td>0.0006 +/- 0.0008</td>\n",
       "      <td>0.0006 +/- 0.0005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>-0.0004 +/- 0.0006</td>\n",
       "      <td>0.0939 +/- 0.0026</td>\n",
       "      <td>-0.0015 +/- 0.0012</td>\n",
       "      <td>0.0409 +/- 0.0015</td>\n",
       "      <td>0.0048 +/- 0.0009</td>\n",
       "      <td>-0.0071 +/- 0.0018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.0224 +/- 0.0010</td>\n",
       "      <td>0.0644 +/- 0.0006</td>\n",
       "      <td>0.0664 +/- 0.0015</td>\n",
       "      <td>0.0449 +/- 0.0011</td>\n",
       "      <td>0.0353 +/- 0.0018</td>\n",
       "      <td>0.0676 +/- 0.0031</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.0003 +/- 0.0004</td>\n",
       "      <td>0.0024 +/- 0.0017</td>\n",
       "      <td>-0.0008 +/- 0.0010</td>\n",
       "      <td>-0.0003 +/- 0.0004</td>\n",
       "      <td>0.0004 +/- 0.0005</td>\n",
       "      <td>-0.0002 +/- 0.0014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>-0.0019 +/- 0.0002</td>\n",
       "      <td>0.0213 +/- 0.0018</td>\n",
       "      <td>0.0073 +/- 0.0007</td>\n",
       "      <td>0.0090 +/- 0.0004</td>\n",
       "      <td>0.0061 +/- 0.0006</td>\n",
       "      <td>0.0038 +/- 0.0002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>-0.0017 +/- 0.0002</td>\n",
       "      <td>0.0126 +/- 0.0012</td>\n",
       "      <td>0.0002 +/- 0.0006</td>\n",
       "      <td>0.0045 +/- 0.0004</td>\n",
       "      <td>0.0026 +/- 0.0003</td>\n",
       "      <td>-0.0005 +/- 0.0002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.0184 +/- 0.0015</td>\n",
       "      <td>0.0140 +/- 0.0020</td>\n",
       "      <td>0.0279 +/- 0.0011</td>\n",
       "      <td>0.0163 +/- 0.0007</td>\n",
       "      <td>0.0123 +/- 0.0004</td>\n",
       "      <td>0.0140 +/- 0.0010</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              svm_poly             svm_rbf              linsvm  \\\n",
       "0   -0.0013 +/- 0.0005   0.0014 +/- 0.0014  -0.0010 +/- 0.0002   \n",
       "1    0.0000 +/- 0.0005   0.0030 +/- 0.0018   0.0003 +/- 0.0005   \n",
       "2    0.0003 +/- 0.0005  -0.0011 +/- 0.0006  -0.0016 +/- 0.0017   \n",
       "3   -0.0000 +/- 0.0003   0.0002 +/- 0.0010  -0.0014 +/- 0.0013   \n",
       "4    0.0000 +/- 0.0004   0.0001 +/- 0.0006  -0.0008 +/- 0.0005   \n",
       "5    0.0001 +/- 0.0003   0.0013 +/- 0.0012  -0.0015 +/- 0.0003   \n",
       "6    0.0031 +/- 0.0005   0.0106 +/- 0.0012   0.0062 +/- 0.0011   \n",
       "7    0.0004 +/- 0.0003   0.0012 +/- 0.0006   0.0001 +/- 0.0008   \n",
       "8   -0.0013 +/- 0.0006   0.0468 +/- 0.0036  -0.0015 +/- 0.0013   \n",
       "9    0.0001 +/- 0.0003   0.0037 +/- 0.0010  -0.0008 +/- 0.0003   \n",
       "10  -0.0004 +/- 0.0006   0.0939 +/- 0.0026  -0.0015 +/- 0.0012   \n",
       "11   0.0224 +/- 0.0010   0.0644 +/- 0.0006   0.0664 +/- 0.0015   \n",
       "12   0.0003 +/- 0.0004   0.0024 +/- 0.0017  -0.0008 +/- 0.0010   \n",
       "13  -0.0019 +/- 0.0002   0.0213 +/- 0.0018   0.0073 +/- 0.0007   \n",
       "14  -0.0017 +/- 0.0002   0.0126 +/- 0.0012   0.0002 +/- 0.0006   \n",
       "15   0.0184 +/- 0.0015   0.0140 +/- 0.0020   0.0279 +/- 0.0011   \n",
       "\n",
       "                 Dtree                RDF              LogReg  \n",
       "0    0.0048 +/- 0.0004  0.0018 +/- 0.0007   0.0000 +/- 0.0002  \n",
       "1    0.0000 +/- 0.0002  0.0008 +/- 0.0005   0.0017 +/- 0.0011  \n",
       "2   -0.0005 +/- 0.0004  0.0020 +/- 0.0005  -0.0001 +/- 0.0009  \n",
       "3   -0.0003 +/- 0.0001  0.0010 +/- 0.0003   0.0007 +/- 0.0015  \n",
       "4    0.0000 +/- 0.0000  0.0000 +/- 0.0001   0.0001 +/- 0.0003  \n",
       "5   -0.0006 +/- 0.0009  0.0005 +/- 0.0007  -0.0010 +/- 0.0004  \n",
       "6    0.0061 +/- 0.0007  0.0022 +/- 0.0006   0.0080 +/- 0.0010  \n",
       "7    0.0002 +/- 0.0001  0.0005 +/- 0.0006   0.0006 +/- 0.0011  \n",
       "8    0.0178 +/- 0.0008  0.0044 +/- 0.0006  -0.0049 +/- 0.0013  \n",
       "9    0.0004 +/- 0.0005  0.0006 +/- 0.0008   0.0006 +/- 0.0005  \n",
       "10   0.0409 +/- 0.0015  0.0048 +/- 0.0009  -0.0071 +/- 0.0018  \n",
       "11   0.0449 +/- 0.0011  0.0353 +/- 0.0018   0.0676 +/- 0.0031  \n",
       "12  -0.0003 +/- 0.0004  0.0004 +/- 0.0005  -0.0002 +/- 0.0014  \n",
       "13   0.0090 +/- 0.0004  0.0061 +/- 0.0006   0.0038 +/- 0.0002  \n",
       "14   0.0045 +/- 0.0004  0.0026 +/- 0.0003  -0.0005 +/- 0.0002  \n",
       "15   0.0163 +/- 0.0007  0.0123 +/- 0.0004   0.0140 +/- 0.0010  "
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "importance_matrix = []\n",
    "for f in range(16):\n",
    "    data = pd.Series({'svm_poly' : f\"{model_r['svm_poly_0.001_0.90_0.89']['importances_mean'][f]:.4f} +/- {model_r['svm_poly_0.001_0.90_0.89']['importances_std'][f]:.4f}\",\n",
    "                    'svm_rbf' : f\"{model_r['svm_rbf_10_0.84_0.86']['importances_mean'][f]:.4f} +/- {model_r['svm_rbf_10_0.84_0.86']['importances_std'][f]:.4f}\",\n",
    "                    'linsvm' : f\"{model_r['svm_linear_10_0.86_0.87']['importances_mean'][f]:.4f} +/- {model_r['svm_linear_10_0.86_0.87']['importances_std'][f]:.4f}\",\n",
    "                    'Dtree' : f\"{model_r['dtree_8_0.90_0.90']['importances_mean'][f]:.4f} +/- {model_r['dtree_8_0.90_0.90']['importances_std'][f]:.4f}\",\n",
    "                    'RDF' : f\"{model_r['rdf_10_500_0.90_0.90']['importances_mean'][f]:.4f} +/- {model_r['rdf_10_500_0.90_0.90']['importances_std'][f]:.4f}\",\n",
    "                    'LogReg' : f\"{model_r['logreg_0.001_0.83_0.85']['importances_mean'][f]:.4f} +/- {model_r['logreg_0.001_0.83_0.85']['importances_std'][f]:.4f}\"})\n",
    "    importance_matrix.append(data)\n",
    "\n",
    "result = pd.DataFrame(importance_matrix)\n",
    "result"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.10.4 64-bit",
   "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": "3.10.4"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "8af0c7b516e23976b6ef6c4f3f67ca7449b209a47950dd9299e28344133359b8"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}