CNN_on_MNIST_Fashion_Dataset.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CNN on MNIST Fashion Dataset.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "ennWYXmnF6nw"
      },
      "source": [
        "#########################################\n",
        "# CNN on MNIST Fashion Dataset\n",
        "# By Giriraj Vaishnav\n",
        "#########################################"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SxkWNyogGDmy"
      },
      "source": [
        "# Importing the necessary libraries\n",
        "\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense\n",
        "from keras.utils import np_utils\n",
        "from keras.datasets import fashion_mnist\n",
        "from keras.layers import Dropout\n",
        "from keras.layers import Flatten\n",
        "from keras.layers.convolutional import Conv2D\n",
        "from keras.layers.convolutional import MaxPooling2D"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YPEb60j6GZaF",
        "outputId": "6dab6321-5200-4b04-9fca-ce58fe744c58"
      },
      "source": [
        "# Set the seed for reproducibility and load the data MNIST dataset\n",
        "\n",
        "seed=10\n",
        "np.random.seed(seed)\n",
        "(X_train,y_train), (X_test, y_test)= fashion_mnist.load_data()"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz\n",
            "32768/29515 [=================================] - 0s 0us/step\n",
            "40960/29515 [=========================================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz\n",
            "26427392/26421880 [==============================] - 0s 0us/step\n",
            "26435584/26421880 [==============================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz\n",
            "16384/5148 [===============================================================================================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz\n",
            "4423680/4422102 [==============================] - 0s 0us/step\n",
            "4431872/4422102 [==============================] - 0s 0us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "f1r2BEBlG4Bj"
      },
      "source": [
        "# Convert the data into float values\n",
        "\n",
        "X_train=X_train.reshape(X_train.shape[0], 1,28,28).astype('float32')\n",
        "X_test=X_test.reshape(X_test.shape[0], 1,28,28).astype('float32')"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0rDUTUwuHAbt",
        "outputId": "50dafd68-d3f3-488a-dfed-a581b1cbb122"
      },
      "source": [
        "# Normalize the data\n",
        "\n",
        "X_train=X_train/255\n",
        "X_test=X_test/255\n",
        "y_train= np_utils.to_categorical(y_train)\n",
        "y_test= np_utils.to_categorical(y_test)\n",
        "num_classes=y_train.shape[1]\n",
        "print(num_classes)"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "10\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9NxZC71RHD5i",
        "outputId": "180a73e0-e3b3-4242-b0a8-55418e70d40d"
      },
      "source": [
        "# Training the CNN model \n",
        "\n",
        "model=Sequential()\n",
        "model.add(Conv2D(32,5,5, padding='same',input_shape=(1,28,28), activation='relu'))\n",
        "model.add(MaxPooling2D(pool_size=(2,2), padding='same'))\n",
        "model.add(Dropout(0.2))\n",
        "model.add(Flatten())\n",
        "model.add(Dense(128, activation='relu'))\n",
        "model.add(Dense(num_classes, activation='softmax'))\n",
        "model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
        "H = model.fit(X_train, y_train, validation_data=(X_test,y_test),epochs=10, batch_size=200, verbose=2)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "300/300 - 17s - loss: 0.8573 - accuracy: 0.6967 - val_loss: 0.5492 - val_accuracy: 0.8019\n",
            "Epoch 2/10\n",
            "300/300 - 2s - loss: 0.5296 - accuracy: 0.8094 - val_loss: 0.4624 - val_accuracy: 0.8339\n",
            "Epoch 3/10\n",
            "300/300 - 2s - loss: 0.4756 - accuracy: 0.8281 - val_loss: 0.4317 - val_accuracy: 0.8460\n",
            "Epoch 4/10\n",
            "300/300 - 2s - loss: 0.4454 - accuracy: 0.8390 - val_loss: 0.4130 - val_accuracy: 0.8508\n",
            "Epoch 5/10\n",
            "300/300 - 2s - loss: 0.4252 - accuracy: 0.8470 - val_loss: 0.4037 - val_accuracy: 0.8542\n",
            "Epoch 6/10\n",
            "300/300 - 2s - loss: 0.4091 - accuracy: 0.8522 - val_loss: 0.3868 - val_accuracy: 0.8606\n",
            "Epoch 7/10\n",
            "300/300 - 2s - loss: 0.4010 - accuracy: 0.8546 - val_loss: 0.3907 - val_accuracy: 0.8595\n",
            "Epoch 8/10\n",
            "300/300 - 2s - loss: 0.3897 - accuracy: 0.8589 - val_loss: 0.3728 - val_accuracy: 0.8646\n",
            "Epoch 9/10\n",
            "300/300 - 2s - loss: 0.3845 - accuracy: 0.8605 - val_loss: 0.3702 - val_accuracy: 0.8679\n",
            "Epoch 10/10\n",
            "300/300 - 2s - loss: 0.3753 - accuracy: 0.8642 - val_loss: 0.3692 - val_accuracy: 0.8646\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ny8yr_AvHYW6",
        "outputId": "925ffba0-fff2-4673-ec04-9fa11c26f367"
      },
      "source": [
        "# Evaluating the model\n",
        "\n",
        "score= model.evaluate(X_test, y_test, verbose=0)\n",
        "print('The error is: %.2f%%'%(100-score[1]*100))"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "The error is: 13.54%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "v39w-PAdHcWO",
        "outputId": "b485b22f-f761-4783-c0ff-2cb470e3cb73"
      },
      "source": [
        "# Ploting the graph for Training and Validation Loss\n",
        "\n",
        "from matplotlib import pyplot as plt\n",
        "plt.plot(range(0,10), H.history[\"loss\"], \"r\", label = \"Training Loss\")\n",
        "plt.plot(range(0,10), H.history[\"val_loss\"], \"b\", label = \"Validation Loss\")\n",
        "plt.title(\"Loss in CNN\")\n",
        "plt.legend(loc = \"upper right\")\n",
        "plt.xlabel(\"Number of Epochs\")\n",
        "plt.ylabel(\"Training and Validation Loss\")\n",
        "plt.show()"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "vjQ56uLoHiPh",
        "outputId": "fc27fe4d-9799-42ce-b8c5-f83c5cb4cc68"
      },
      "source": [
        "# Ploting the graph for Training and Validation Accuracy\n",
        "\n",
        "from matplotlib import pyplot as plt\n",
        "plt.plot(range(0,10), H.history[\"accuracy\"], \"r\", label = \"Training Accuracy\")\n",
        "plt.plot(range(0,10), H.history[\"val_accuracy\"], \"b\", label = \"Validation Accuracy\")\n",
        "plt.title(\"Accuracy in CNN\")\n",
        "plt.legend(loc = \"lower right\")\n",
        "plt.xlabel(\"Number of Epochs\")\n",
        "plt.ylabel(\"Training and Validation Accuracy\")\n",
        "plt.show()"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JNJTYgIoHlNe",
        "outputId": "ec31e0c0-a913-40b4-ef1c-ed3aecc0ad80"
      },
      "source": [
        "# Predicting the Test Set\n",
        "\n",
        "pred_labels = model.predict_classes(X_test)\n",
        "print(pred_labels)\n",
        "test_label = np.argmax(y_test, axis=1)\n",
        "print(test_label)"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/keras/engine/sequential.py:450: UserWarning: `model.predict_classes()` is deprecated and will be removed after 2021-01-01. Please use instead:* `np.argmax(model.predict(x), axis=-1)`,   if your model does multi-class classification   (e.g. if it uses a `softmax` last-layer activation).* `(model.predict(x) > 0.5).astype(\"int32\")`,   if your model does binary classification   (e.g. if it uses a `sigmoid` last-layer activation).\n",
            "  warnings.warn('`model.predict_classes()` is deprecated and '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "[9 2 1 ... 8 1 5]\n",
            "[9 2 1 ... 8 1 5]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aBYnvYcdHnd_",
        "outputId": "d6383b97-cf1d-46d5-dc51-784373ab7b8a"
      },
      "source": [
        "# Confusion matrix\n",
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "\n",
        "matrix = confusion_matrix(test_label, pred_labels)\n",
        "print('Confusion matrix : \\n',matrix)"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Confusion matrix : \n",
            " [[843   1  11  33   6   2  93   0  11   0]\n",
            " [  4 956   0  30   3   0   7   0   0   0]\n",
            " [ 14   0 758  12 118   0  97   0   1   0]\n",
            " [ 39   7  13 871  37   0  30   0   3   0]\n",
            " [  0   0  87  40 796   0  76   0   1   0]\n",
            " [  1   0   0   1   0 963   0  18   2  15]\n",
            " [165   1  87  30  89   0 619   0   9   0]\n",
            " [  0   0   0   0   0  30   0 929   1  40]\n",
            " [  3   1   4   7   5   3  15   5 957   0]\n",
            " [  1   0   0   0   0   8   0  37   0 954]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aIM-ouKmHqh3",
        "outputId": "d5b97654-550f-45d8-d922-674939b9afdc"
      },
      "source": [
        "# Classification report for precision, recall f1-score and accuracy\n",
        "\n",
        "from sklearn.metrics import classification_report\n",
        "\n",
        "report = classification_report(test_label,pred_labels)\n",
        "print('Classification report : \\n',report)"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Classification report : \n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "           0       0.79      0.84      0.81      1000\n",
            "           1       0.99      0.96      0.97      1000\n",
            "           2       0.79      0.76      0.77      1000\n",
            "           3       0.85      0.87      0.86      1000\n",
            "           4       0.76      0.80      0.78      1000\n",
            "           5       0.96      0.96      0.96      1000\n",
            "           6       0.66      0.62      0.64      1000\n",
            "           7       0.94      0.93      0.93      1000\n",
            "           8       0.97      0.96      0.96      1000\n",
            "           9       0.95      0.95      0.95      1000\n",
            "\n",
            "    accuracy                           0.86     10000\n",
            "   macro avg       0.86      0.86      0.86     10000\n",
            "weighted avg       0.86      0.86      0.86     10000\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}