CNN_brain_team.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CNN_brain_team.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "3CpXAlETkVnw"
      },
      "source": [
        "import tensorflow as tf\n",
        "import sys\n",
        "from matplotlib import pyplot\n",
        "from keras.datasets import cifar10\n",
        "from tensorflow.keras.utils import to_categorical\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Conv2D\n",
        "from keras.layers import MaxPooling2D\n",
        "from keras.layers import Dense, Dropout\n",
        "from keras.layers import Flatten\n",
        "from keras.optimizers import SGD\n",
        "from tensorflow.keras import datasets, layers, models\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "q80ViD-FrD25"
      },
      "source": [
        "[-1, 0 , 15]\n",
        "34 \n",
        "35 "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VHDrmVaqj6w5"
      },
      "source": [
        "# Using CIFAR-10 dataset, change train and test vars appropriately \n",
        "\n",
        "def load_dataset():\n",
        "\t# load dataset\n",
        "\t(trainX, trainY), (testX, testY) = cifar10.load_data()\n",
        "\t# one hot encode target values\n",
        "\ttrainY = to_categorical(trainY)\n",
        "\ttestY = to_categorical(testY)\n",
        "\treturn trainX, trainY, testX, testY\n",
        "# scale pixels\n",
        "def prep_pixels(train, test):\n",
        "\t# convert from integers to floats\n",
        "\ttrain_norm = train.astype('float32')\n",
        "\ttest_norm = test.astype('float32')\n",
        "\t# normalize to range 0-1\n",
        "\ttrain_norm = train_norm / 255.0\n",
        "\ttest_norm = test_norm / 255.0\n",
        "\t# return normalized images\n",
        "\treturn train_norm, test_norm\n",
        "\n",
        "#plot metrics\n",
        "def Loss_ACC_plot(history):\n",
        "\t# plot loss\n",
        "    plt.figure()\n",
        "    plt.subplot(1,2,1)\n",
        "    plt.title('Cross Entropy Loss')\n",
        "    plt.plot(history.history['loss'], color='blue', label='train')\n",
        "    plt.plot(history.history['val_loss'], color='orange', label='validation')\n",
        "    # plot accuracy\n",
        "    plt.subplot(1,2,2)\n",
        "    plt.title('Classification Accuracy')\n",
        "    plt.plot(history.history['accuracy'], color='blue', label='train')\n",
        "    plt.plot(history.history['val_accuracy'], color='orange', label='validation')\n",
        " \n",
        "# define cnn model\n",
        "def define_model():\n",
        "\tmodel = Sequential()\n",
        "\tmodel.add(Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same', input_shape=(32, 32, 3)))\n",
        "\tmodel.add(Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same'))\n",
        "\tmodel.add(MaxPooling2D((2, 2)))\n",
        "\tmodel.add(Dropout(0.2))\n",
        "\tmodel.add(Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same'))\n",
        "\tmodel.add(Conv2D(64, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same'))\n",
        "\tmodel.add(MaxPooling2D((2, 2)))\n",
        "\tmodel.add(Dropout(0.2))\n",
        "\tmodel.add(Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same'))\n",
        "\tmodel.add(Conv2D(128, (3, 3), activation='relu', kernel_initializer='he_uniform', padding='same'))\n",
        "\tmodel.add(MaxPooling2D((2, 2)))\n",
        "\tmodel.add(Dropout(0.2))\n",
        "\tmodel.add(Flatten())\n",
        "\tmodel.add(Dense(128, activation='relu', kernel_initializer='he_uniform'))\n",
        "\tmodel.add(Dropout(0.2))\n",
        "\tmodel.add(Dense(10, activation='softmax'))\n",
        "\t# compile model\n",
        "\topt = SGD(learning_rate=0.001, momentum=0.9)\n",
        "\tmodel.compile(optimizer=opt, loss='categorical_crossentropy', metrics=['accuracy'])\n",
        "\treturn model"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "De1BS7y5fOIl",
        "outputId": "d2c15b8f-5f79-481c-a946-0cd8aa6eafac"
      },
      "source": [
        "# -*- coding: utf-8 -*-\n",
        "\"\"\"\n",
        "Created on Wed Jun 23 11:27:26 2021\n",
        "@author: Pablo\n",
        "\"\"\"\n",
        "\n",
        "#Ensure that we are using GPU\n",
        "gpus = tf.config.experimental.list_physical_devices('GPU')\n",
        "if gpus:\n",
        "    try:\n",
        "        for gpu in gpus:\n",
        "            tf.config.experimental.set_memory_growth(gpu, True)\n",
        "            print(tf.test.gpu_device_name())\n",
        "    except RuntimeError as e:\n",
        "        print(e)\n",
        "\n",
        "\n",
        "x_train, y_train, x_test, y_test = load_dataset()\n",
        "# prepare pixel data\n",
        "x_train, x_test = prep_pixels(x_train, x_test)\n",
        "# define model\n",
        "model = define_model()\n",
        "# fit model\n",
        "history = model.fit(x_train, y_train,validation_split=.2, epochs=100,shuffle = True, batch_size=64)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/device:GPU:0\n",
            "Epoch 1/100\n",
            "625/625 [==============================] - 7s 9ms/step - loss: 2.2180 - accuracy: 0.1708 - val_loss: 1.8350 - val_accuracy: 0.3333\n",
            "Epoch 2/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 1.7969 - accuracy: 0.3389 - val_loss: 1.5684 - val_accuracy: 0.4272\n",
            "Epoch 3/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 1.6192 - accuracy: 0.4021 - val_loss: 1.5209 - val_accuracy: 0.4508\n",
            "Epoch 4/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 1.5313 - accuracy: 0.4381 - val_loss: 1.4168 - val_accuracy: 0.4904\n",
            "Epoch 5/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 1.4487 - accuracy: 0.4702 - val_loss: 1.3263 - val_accuracy: 0.5311\n",
            "Epoch 6/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 1.3820 - accuracy: 0.4959 - val_loss: 1.2817 - val_accuracy: 0.5380\n",
            "Epoch 7/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 1.3255 - accuracy: 0.5197 - val_loss: 1.2044 - val_accuracy: 0.5710\n",
            "Epoch 8/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 1.2714 - accuracy: 0.5412 - val_loss: 1.1570 - val_accuracy: 0.5914\n",
            "Epoch 9/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 1.2267 - accuracy: 0.5581 - val_loss: 1.1454 - val_accuracy: 0.5925\n",
            "Epoch 10/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 1.1844 - accuracy: 0.5751 - val_loss: 1.1141 - val_accuracy: 0.6040\n",
            "Epoch 11/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 1.1394 - accuracy: 0.5905 - val_loss: 1.0524 - val_accuracy: 0.6268\n",
            "Epoch 12/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 1.1076 - accuracy: 0.5997 - val_loss: 0.9941 - val_accuracy: 0.6505\n",
            "Epoch 13/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 1.0580 - accuracy: 0.6196 - val_loss: 0.9765 - val_accuracy: 0.6518\n",
            "Epoch 14/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 1.0370 - accuracy: 0.6287 - val_loss: 0.9322 - val_accuracy: 0.6698\n",
            "Epoch 15/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.9972 - accuracy: 0.6456 - val_loss: 0.9245 - val_accuracy: 0.6722\n",
            "Epoch 16/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.9785 - accuracy: 0.6479 - val_loss: 0.8887 - val_accuracy: 0.6862\n",
            "Epoch 17/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.9564 - accuracy: 0.6635 - val_loss: 0.8823 - val_accuracy: 0.6892\n",
            "Epoch 18/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.9283 - accuracy: 0.6681 - val_loss: 0.9067 - val_accuracy: 0.6779\n",
            "Epoch 19/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.9044 - accuracy: 0.6764 - val_loss: 0.8443 - val_accuracy: 0.7015\n",
            "Epoch 20/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.8829 - accuracy: 0.6843 - val_loss: 0.8436 - val_accuracy: 0.7037\n",
            "Epoch 21/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.8665 - accuracy: 0.6902 - val_loss: 0.8185 - val_accuracy: 0.7130\n",
            "Epoch 22/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.8491 - accuracy: 0.6986 - val_loss: 0.8186 - val_accuracy: 0.7140\n",
            "Epoch 23/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.8287 - accuracy: 0.7079 - val_loss: 0.7762 - val_accuracy: 0.7265\n",
            "Epoch 24/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.8116 - accuracy: 0.7090 - val_loss: 0.7498 - val_accuracy: 0.7395\n",
            "Epoch 25/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.7915 - accuracy: 0.7179 - val_loss: 0.7563 - val_accuracy: 0.7353\n",
            "Epoch 26/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.7724 - accuracy: 0.7262 - val_loss: 0.7417 - val_accuracy: 0.7397\n",
            "Epoch 27/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.7584 - accuracy: 0.7331 - val_loss: 0.7610 - val_accuracy: 0.7347\n",
            "Epoch 28/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.7432 - accuracy: 0.7381 - val_loss: 0.7313 - val_accuracy: 0.7487\n",
            "Epoch 29/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.7282 - accuracy: 0.7416 - val_loss: 0.7239 - val_accuracy: 0.7477\n",
            "Epoch 30/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.7211 - accuracy: 0.7408 - val_loss: 0.7290 - val_accuracy: 0.7444\n",
            "Epoch 31/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.7042 - accuracy: 0.7510 - val_loss: 0.6926 - val_accuracy: 0.7607\n",
            "Epoch 32/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.6890 - accuracy: 0.7544 - val_loss: 0.6976 - val_accuracy: 0.7572\n",
            "Epoch 33/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.6759 - accuracy: 0.7599 - val_loss: 0.6777 - val_accuracy: 0.7656\n",
            "Epoch 34/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.6763 - accuracy: 0.7604 - val_loss: 0.6805 - val_accuracy: 0.7682\n",
            "Epoch 35/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.6623 - accuracy: 0.7662 - val_loss: 0.6643 - val_accuracy: 0.7696\n",
            "Epoch 36/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.6395 - accuracy: 0.7728 - val_loss: 0.6758 - val_accuracy: 0.7676\n",
            "Epoch 37/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.6285 - accuracy: 0.7774 - val_loss: 0.6822 - val_accuracy: 0.7646\n",
            "Epoch 38/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.6270 - accuracy: 0.7791 - val_loss: 0.6565 - val_accuracy: 0.7734\n",
            "Epoch 39/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.6218 - accuracy: 0.7798 - val_loss: 0.6623 - val_accuracy: 0.7721\n",
            "Epoch 40/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.5986 - accuracy: 0.7869 - val_loss: 0.6387 - val_accuracy: 0.7800\n",
            "Epoch 41/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.6026 - accuracy: 0.7884 - val_loss: 0.6366 - val_accuracy: 0.7813\n",
            "Epoch 42/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.5843 - accuracy: 0.7867 - val_loss: 0.6428 - val_accuracy: 0.7789\n",
            "Epoch 43/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.5733 - accuracy: 0.7969 - val_loss: 0.6545 - val_accuracy: 0.7765\n",
            "Epoch 44/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.5717 - accuracy: 0.7991 - val_loss: 0.6414 - val_accuracy: 0.7798\n",
            "Epoch 45/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.5579 - accuracy: 0.8004 - val_loss: 0.6371 - val_accuracy: 0.7826\n",
            "Epoch 46/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.5476 - accuracy: 0.8025 - val_loss: 0.6363 - val_accuracy: 0.7822\n",
            "Epoch 47/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.5366 - accuracy: 0.8116 - val_loss: 0.6394 - val_accuracy: 0.7830\n",
            "Epoch 48/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.5347 - accuracy: 0.8122 - val_loss: 0.6109 - val_accuracy: 0.7908\n",
            "Epoch 49/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.5197 - accuracy: 0.8122 - val_loss: 0.6196 - val_accuracy: 0.7883\n",
            "Epoch 50/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.5136 - accuracy: 0.8167 - val_loss: 0.6177 - val_accuracy: 0.7899\n",
            "Epoch 51/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.5029 - accuracy: 0.8222 - val_loss: 0.5953 - val_accuracy: 0.7986\n",
            "Epoch 52/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.5053 - accuracy: 0.8208 - val_loss: 0.6036 - val_accuracy: 0.7941\n",
            "Epoch 53/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4908 - accuracy: 0.8233 - val_loss: 0.6117 - val_accuracy: 0.7951\n",
            "Epoch 54/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4870 - accuracy: 0.8254 - val_loss: 0.6004 - val_accuracy: 0.7979\n",
            "Epoch 55/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4793 - accuracy: 0.8284 - val_loss: 0.6169 - val_accuracy: 0.7945\n",
            "Epoch 56/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4792 - accuracy: 0.8314 - val_loss: 0.5919 - val_accuracy: 0.8026\n",
            "Epoch 57/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4626 - accuracy: 0.8357 - val_loss: 0.6244 - val_accuracy: 0.7933\n",
            "Epoch 58/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4659 - accuracy: 0.8315 - val_loss: 0.5958 - val_accuracy: 0.8007\n",
            "Epoch 59/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4603 - accuracy: 0.8374 - val_loss: 0.5941 - val_accuracy: 0.8001\n",
            "Epoch 60/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4506 - accuracy: 0.8399 - val_loss: 0.5858 - val_accuracy: 0.8033\n",
            "Epoch 61/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4382 - accuracy: 0.8414 - val_loss: 0.6071 - val_accuracy: 0.7958\n",
            "Epoch 62/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4286 - accuracy: 0.8474 - val_loss: 0.5996 - val_accuracy: 0.8015\n",
            "Epoch 63/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4256 - accuracy: 0.8491 - val_loss: 0.5921 - val_accuracy: 0.8051\n",
            "Epoch 64/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4242 - accuracy: 0.8467 - val_loss: 0.6240 - val_accuracy: 0.7931\n",
            "Epoch 65/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4124 - accuracy: 0.8543 - val_loss: 0.5881 - val_accuracy: 0.8063\n",
            "Epoch 66/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4156 - accuracy: 0.8513 - val_loss: 0.5918 - val_accuracy: 0.8071\n",
            "Epoch 67/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.4042 - accuracy: 0.8553 - val_loss: 0.5969 - val_accuracy: 0.8031\n",
            "Epoch 68/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3919 - accuracy: 0.8581 - val_loss: 0.5950 - val_accuracy: 0.8041\n",
            "Epoch 69/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3973 - accuracy: 0.8591 - val_loss: 0.5937 - val_accuracy: 0.8023\n",
            "Epoch 70/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3927 - accuracy: 0.8578 - val_loss: 0.5814 - val_accuracy: 0.8109\n",
            "Epoch 71/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3806 - accuracy: 0.8631 - val_loss: 0.5849 - val_accuracy: 0.8113\n",
            "Epoch 72/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3804 - accuracy: 0.8646 - val_loss: 0.5889 - val_accuracy: 0.8135\n",
            "Epoch 73/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3664 - accuracy: 0.8700 - val_loss: 0.5957 - val_accuracy: 0.8056\n",
            "Epoch 74/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3647 - accuracy: 0.8688 - val_loss: 0.5934 - val_accuracy: 0.8082\n",
            "Epoch 75/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3594 - accuracy: 0.8689 - val_loss: 0.6028 - val_accuracy: 0.8109\n",
            "Epoch 76/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3527 - accuracy: 0.8717 - val_loss: 0.5965 - val_accuracy: 0.8081\n",
            "Epoch 77/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3574 - accuracy: 0.8722 - val_loss: 0.5916 - val_accuracy: 0.8123\n",
            "Epoch 78/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3441 - accuracy: 0.8775 - val_loss: 0.6023 - val_accuracy: 0.8104\n",
            "Epoch 79/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3418 - accuracy: 0.8762 - val_loss: 0.5842 - val_accuracy: 0.8135\n",
            "Epoch 80/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3379 - accuracy: 0.8778 - val_loss: 0.5979 - val_accuracy: 0.8149\n",
            "Epoch 81/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3299 - accuracy: 0.8811 - val_loss: 0.5981 - val_accuracy: 0.8122\n",
            "Epoch 82/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3261 - accuracy: 0.8819 - val_loss: 0.6144 - val_accuracy: 0.8093\n",
            "Epoch 83/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3114 - accuracy: 0.8876 - val_loss: 0.6163 - val_accuracy: 0.8064\n",
            "Epoch 84/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3176 - accuracy: 0.8867 - val_loss: 0.5852 - val_accuracy: 0.8186\n",
            "Epoch 85/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3164 - accuracy: 0.8864 - val_loss: 0.5993 - val_accuracy: 0.8099\n",
            "Epoch 86/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3078 - accuracy: 0.8892 - val_loss: 0.6049 - val_accuracy: 0.8127\n",
            "Epoch 87/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.3064 - accuracy: 0.8913 - val_loss: 0.6140 - val_accuracy: 0.8131\n",
            "Epoch 88/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.2984 - accuracy: 0.8928 - val_loss: 0.6014 - val_accuracy: 0.8155\n",
            "Epoch 89/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.2953 - accuracy: 0.8926 - val_loss: 0.6364 - val_accuracy: 0.8102\n",
            "Epoch 90/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.2933 - accuracy: 0.8922 - val_loss: 0.6029 - val_accuracy: 0.8171\n",
            "Epoch 91/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.2817 - accuracy: 0.8975 - val_loss: 0.6020 - val_accuracy: 0.8162\n",
            "Epoch 92/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.2826 - accuracy: 0.8963 - val_loss: 0.6236 - val_accuracy: 0.8108\n",
            "Epoch 93/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.2774 - accuracy: 0.8992 - val_loss: 0.6131 - val_accuracy: 0.8131\n",
            "Epoch 94/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.2784 - accuracy: 0.8994 - val_loss: 0.6049 - val_accuracy: 0.8129\n",
            "Epoch 95/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.2657 - accuracy: 0.9037 - val_loss: 0.6250 - val_accuracy: 0.8156\n",
            "Epoch 96/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.2740 - accuracy: 0.9012 - val_loss: 0.6250 - val_accuracy: 0.8138\n",
            "Epoch 97/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.2617 - accuracy: 0.9057 - val_loss: 0.6142 - val_accuracy: 0.8146\n",
            "Epoch 98/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.2619 - accuracy: 0.9051 - val_loss: 0.6159 - val_accuracy: 0.8168\n",
            "Epoch 99/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.2591 - accuracy: 0.9069 - val_loss: 0.6215 - val_accuracy: 0.8119\n",
            "Epoch 100/100\n",
            "625/625 [==============================] - 5s 8ms/step - loss: 0.2513 - accuracy: 0.9088 - val_loss: 0.6589 - val_accuracy: 0.8140\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        },
        "id": "SGcRPWT7ig5z",
        "outputId": "a48cdf9b-b58a-4fa4-acc3-c45d4e844847"
      },
      "source": [
        "Loss_ACC_plot(history)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SGgttFd_jOXJ",
        "outputId": "541f81a4-6bcf-49d5-aefb-f3e1b137168a"
      },
      "source": [
        "import numpy as np\n",
        "y_pred = model.predict(x_test)\n",
        "y_pred= np.argmax(y_pred, axis=1)\n",
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.metrics import accuracy_score\n",
        "y_test = np.argmax(y_test, axis=1)\n",
        "cmatrix = confusion_matrix(y_test, y_pred)\n",
        "print(cmatrix)\n",
        "ACC = np.trace(cmatrix)/np.sum(cmatrix)\n",
        "print('Testing Acc: ', ACC)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[786  12  45  20  16   3  13  11  61  33]\n",
            " [  9 878   2   0   2   4   8   1  21  75]\n",
            " [ 34   2 709  39  68  43  72  13  11   9]\n",
            " [ 12   5  42 629  73 115  74  22  11  17]\n",
            " [  7   2  32  33 828  13  45  34   5   1]\n",
            " [  4   2  15 154  57 707  29  24   4   4]\n",
            " [  2   1  30  34  12   9 902   3   4   3]\n",
            " [  4   0  19  39  49  34  10 833   3   9]\n",
            " [ 23   8   7  14   6   1   6   4 911  20]\n",
            " [ 14  39   1   9   3   2   3   3  19 907]]\n",
            "Testing Acc:  0.809\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}