fashion_mnist.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "fashion_mnist.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/mammadmaheri7/fashion-mnist/blob/master/fashion_mnist.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1MDgMTHLX-Ap",
        "colab_type": "text"
      },
      "source": [
        "Loading tensorflow and keras and other appropriate libraries:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BKz6EMjUX8Kw",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow import keras\n",
        "\n",
        "import numpy as np\n",
        "\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BGPIVJNnYEER",
        "colab_type": "text"
      },
      "source": [
        "Loading Fashion MNIST dataset from keras datasets and separate test and train datas:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PTwr1KYCYHqK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "(train_images, train_labels),(test_images,test_labels)=keras.datasets.fashion_mnist.load_data()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zCjMKMd9grSB",
        "colab_type": "text"
      },
      "source": [
        "Define the class names of the dataset labels:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PIlEcDmMgrcM",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class_names = ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat',\n",
        "               'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eLHMuBCkgUl3",
        "colab_type": "text"
      },
      "source": [
        "Ploting the first 20 images along with their labels:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LCXH4D7LgfYp",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 482
        },
        "outputId": "93370de0-a3ef-4980-d45c-20614342987b"
      },
      "source": [
        "plt.figure(figsize=(10,10))\n",
        "for i in range(0,20):\n",
        "    plt.subplot(5,5, i+1)\n",
        "    plt.imshow(train_images[i])\n",
        "    plt.title(class_names[(train_labels[i])])\n",
        "    plt.xticks([])\n",
        "    plt.yticks([])\n",
        "plt.show()"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 20 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "C7NndtUmYKRq",
        "colab_type": "text"
      },
      "source": [
        "Printing the shape of the train and test images:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OQLoQbrGYOjL",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "14b1abd8-be7f-4f83-8ed5-dabd655c8969"
      },
      "source": [
        "print(train_images.shape)\n",
        "print(test_images.shape)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(60000, 28, 28)\n",
            "(10000, 28, 28)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Mfv9u8aZYsxa",
        "colab_type": "text"
      },
      "source": [
        "Plot an image figure of the train images for better illustration:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HsKfpH9mY1uB",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "outputId": "0900d7b3-3be8-4c64-fecf-16d17c937a63"
      },
      "source": [
        "plt.figure()\n",
        "plt.imshow(train_images[0],cmap=plt.cm.binary)\n",
        "plt.colorbar()\n",
        "plt.grid(False)"
      ],
      "execution_count": 13,
      "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": "markdown",
      "metadata": {
        "id": "fIZj_27iaL7V",
        "colab_type": "text"
      },
      "source": [
        "The original matrix of an image in out dataset:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2P_8zdzxaDCX",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "95cbf37f-bb11-4ae5-fa5c-1cb6f1ce6860"
      },
      "source": [
        "train_images[0]"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "          0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "          0,   0],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "          0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "          0,   0],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "          0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "          0,   0],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   1,\n",
              "          0,   0,  13,  73,   0,   0,   1,   4,   0,   0,   0,   0,   1,\n",
              "          1,   0],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   3,\n",
              "          0,  36, 136, 127,  62,  54,   0,   0,   0,   1,   3,   4,   0,\n",
              "          0,   3],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   6,\n",
              "          0, 102, 204, 176, 134, 144, 123,  23,   0,   0,   0,   0,  12,\n",
              "         10,   0],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "          0, 155, 236, 207, 178, 107, 156, 161, 109,  64,  23,  77, 130,\n",
              "         72,  15],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   1,   0,\n",
              "         69, 207, 223, 218, 216, 216, 163, 127, 121, 122, 146, 141,  88,\n",
              "        172,  66],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   1,   1,   1,   0,\n",
              "        200, 232, 232, 233, 229, 223, 223, 215, 213, 164, 127, 123, 196,\n",
              "        229,   0],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "        183, 225, 216, 223, 228, 235, 227, 224, 222, 224, 221, 223, 245,\n",
              "        173,   0],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "        193, 228, 218, 213, 198, 180, 212, 210, 211, 213, 223, 220, 243,\n",
              "        202,   0],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   1,   3,   0,  12,\n",
              "        219, 220, 212, 218, 192, 169, 227, 208, 218, 224, 212, 226, 197,\n",
              "        209,  52],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   6,   0,  99,\n",
              "        244, 222, 220, 218, 203, 198, 221, 215, 213, 222, 220, 245, 119,\n",
              "        167,  56],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   4,   0,   0,  55,\n",
              "        236, 228, 230, 228, 240, 232, 213, 218, 223, 234, 217, 217, 209,\n",
              "         92,   0],\n",
              "       [  0,   0,   1,   4,   6,   7,   2,   0,   0,   0,   0,   0, 237,\n",
              "        226, 217, 223, 222, 219, 222, 221, 216, 223, 229, 215, 218, 255,\n",
              "         77,   0],\n",
              "       [  0,   3,   0,   0,   0,   0,   0,   0,   0,  62, 145, 204, 228,\n",
              "        207, 213, 221, 218, 208, 211, 218, 224, 223, 219, 215, 224, 244,\n",
              "        159,   0],\n",
              "       [  0,   0,   0,   0,  18,  44,  82, 107, 189, 228, 220, 222, 217,\n",
              "        226, 200, 205, 211, 230, 224, 234, 176, 188, 250, 248, 233, 238,\n",
              "        215,   0],\n",
              "       [  0,  57, 187, 208, 224, 221, 224, 208, 204, 214, 208, 209, 200,\n",
              "        159, 245, 193, 206, 223, 255, 255, 221, 234, 221, 211, 220, 232,\n",
              "        246,   0],\n",
              "       [  3, 202, 228, 224, 221, 211, 211, 214, 205, 205, 205, 220, 240,\n",
              "         80, 150, 255, 229, 221, 188, 154, 191, 210, 204, 209, 222, 228,\n",
              "        225,   0],\n",
              "       [ 98, 233, 198, 210, 222, 229, 229, 234, 249, 220, 194, 215, 217,\n",
              "        241,  65,  73, 106, 117, 168, 219, 221, 215, 217, 223, 223, 224,\n",
              "        229,  29],\n",
              "       [ 75, 204, 212, 204, 193, 205, 211, 225, 216, 185, 197, 206, 198,\n",
              "        213, 240, 195, 227, 245, 239, 223, 218, 212, 209, 222, 220, 221,\n",
              "        230,  67],\n",
              "       [ 48, 203, 183, 194, 213, 197, 185, 190, 194, 192, 202, 214, 219,\n",
              "        221, 220, 236, 225, 216, 199, 206, 186, 181, 177, 172, 181, 205,\n",
              "        206, 115],\n",
              "       [  0, 122, 219, 193, 179, 171, 183, 196, 204, 210, 213, 207, 211,\n",
              "        210, 200, 196, 194, 191, 195, 191, 198, 192, 176, 156, 167, 177,\n",
              "        210,  92],\n",
              "       [  0,   0,  74, 189, 212, 191, 175, 172, 175, 181, 185, 188, 189,\n",
              "        188, 193, 198, 204, 209, 210, 210, 211, 188, 188, 194, 192, 216,\n",
              "        170,   0],\n",
              "       [  2,   0,   0,   0,  66, 200, 222, 237, 239, 242, 246, 243, 244,\n",
              "        221, 220, 193, 191, 179, 182, 182, 181, 176, 166, 168,  99,  58,\n",
              "          0,   0],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,  40,  61,  44,  72,  41,  35,\n",
              "          0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "          0,   0],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "          0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "          0,   0],\n",
              "       [  0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "          0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,   0,\n",
              "          0,   0]], dtype=uint8)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6IrAOu-ZbzYa",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "76132ec1-ea76-493b-a8eb-b936940458d1"
      },
      "source": [
        "train_images[0].min()"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9beJtjzobS4U",
        "colab_type": "text"
      },
      "source": [
        "Scaling the images of the dataset to a range of 0 to 1:\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QPwhd-QEbdvw",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_images = train_images/255.0\n",
        "test_images = test_images/255.0"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ea_TG0zPbnIo",
        "colab_type": "text"
      },
      "source": [
        "Chacking it again:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "q8KwwsntblZi",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "bccae731-175c-44a6-8b57-f68d2bd14cf7"
      },
      "source": [
        "train_images[0]"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        ],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        ],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        ],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.00392157, 0.        , 0.        ,\n",
              "        0.05098039, 0.28627451, 0.        , 0.        , 0.00392157,\n",
              "        0.01568627, 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.00392157, 0.00392157, 0.        ],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.01176471, 0.        , 0.14117647,\n",
              "        0.53333333, 0.49803922, 0.24313725, 0.21176471, 0.        ,\n",
              "        0.        , 0.        , 0.00392157, 0.01176471, 0.01568627,\n",
              "        0.        , 0.        , 0.01176471],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.02352941, 0.        , 0.4       ,\n",
              "        0.8       , 0.69019608, 0.5254902 , 0.56470588, 0.48235294,\n",
              "        0.09019608, 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.04705882, 0.03921569, 0.        ],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.60784314,\n",
              "        0.9254902 , 0.81176471, 0.69803922, 0.41960784, 0.61176471,\n",
              "        0.63137255, 0.42745098, 0.25098039, 0.09019608, 0.30196078,\n",
              "        0.50980392, 0.28235294, 0.05882353],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.00392157, 0.        , 0.27058824, 0.81176471,\n",
              "        0.8745098 , 0.85490196, 0.84705882, 0.84705882, 0.63921569,\n",
              "        0.49803922, 0.4745098 , 0.47843137, 0.57254902, 0.55294118,\n",
              "        0.34509804, 0.6745098 , 0.25882353],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.00392157,\n",
              "        0.00392157, 0.00392157, 0.        , 0.78431373, 0.90980392,\n",
              "        0.90980392, 0.91372549, 0.89803922, 0.8745098 , 0.8745098 ,\n",
              "        0.84313725, 0.83529412, 0.64313725, 0.49803922, 0.48235294,\n",
              "        0.76862745, 0.89803922, 0.        ],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.71764706, 0.88235294,\n",
              "        0.84705882, 0.8745098 , 0.89411765, 0.92156863, 0.89019608,\n",
              "        0.87843137, 0.87058824, 0.87843137, 0.86666667, 0.8745098 ,\n",
              "        0.96078431, 0.67843137, 0.        ],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.75686275, 0.89411765,\n",
              "        0.85490196, 0.83529412, 0.77647059, 0.70588235, 0.83137255,\n",
              "        0.82352941, 0.82745098, 0.83529412, 0.8745098 , 0.8627451 ,\n",
              "        0.95294118, 0.79215686, 0.        ],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.00392157,\n",
              "        0.01176471, 0.        , 0.04705882, 0.85882353, 0.8627451 ,\n",
              "        0.83137255, 0.85490196, 0.75294118, 0.6627451 , 0.89019608,\n",
              "        0.81568627, 0.85490196, 0.87843137, 0.83137255, 0.88627451,\n",
              "        0.77254902, 0.81960784, 0.20392157],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.02352941, 0.        , 0.38823529, 0.95686275, 0.87058824,\n",
              "        0.8627451 , 0.85490196, 0.79607843, 0.77647059, 0.86666667,\n",
              "        0.84313725, 0.83529412, 0.87058824, 0.8627451 , 0.96078431,\n",
              "        0.46666667, 0.65490196, 0.21960784],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.01568627,\n",
              "        0.        , 0.        , 0.21568627, 0.9254902 , 0.89411765,\n",
              "        0.90196078, 0.89411765, 0.94117647, 0.90980392, 0.83529412,\n",
              "        0.85490196, 0.8745098 , 0.91764706, 0.85098039, 0.85098039,\n",
              "        0.81960784, 0.36078431, 0.        ],\n",
              "       [0.        , 0.        , 0.00392157, 0.01568627, 0.02352941,\n",
              "        0.02745098, 0.00784314, 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.92941176, 0.88627451, 0.85098039,\n",
              "        0.8745098 , 0.87058824, 0.85882353, 0.87058824, 0.86666667,\n",
              "        0.84705882, 0.8745098 , 0.89803922, 0.84313725, 0.85490196,\n",
              "        1.        , 0.30196078, 0.        ],\n",
              "       [0.        , 0.01176471, 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.24313725,\n",
              "        0.56862745, 0.8       , 0.89411765, 0.81176471, 0.83529412,\n",
              "        0.86666667, 0.85490196, 0.81568627, 0.82745098, 0.85490196,\n",
              "        0.87843137, 0.8745098 , 0.85882353, 0.84313725, 0.87843137,\n",
              "        0.95686275, 0.62352941, 0.        ],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.07058824,\n",
              "        0.17254902, 0.32156863, 0.41960784, 0.74117647, 0.89411765,\n",
              "        0.8627451 , 0.87058824, 0.85098039, 0.88627451, 0.78431373,\n",
              "        0.80392157, 0.82745098, 0.90196078, 0.87843137, 0.91764706,\n",
              "        0.69019608, 0.7372549 , 0.98039216, 0.97254902, 0.91372549,\n",
              "        0.93333333, 0.84313725, 0.        ],\n",
              "       [0.        , 0.22352941, 0.73333333, 0.81568627, 0.87843137,\n",
              "        0.86666667, 0.87843137, 0.81568627, 0.8       , 0.83921569,\n",
              "        0.81568627, 0.81960784, 0.78431373, 0.62352941, 0.96078431,\n",
              "        0.75686275, 0.80784314, 0.8745098 , 1.        , 1.        ,\n",
              "        0.86666667, 0.91764706, 0.86666667, 0.82745098, 0.8627451 ,\n",
              "        0.90980392, 0.96470588, 0.        ],\n",
              "       [0.01176471, 0.79215686, 0.89411765, 0.87843137, 0.86666667,\n",
              "        0.82745098, 0.82745098, 0.83921569, 0.80392157, 0.80392157,\n",
              "        0.80392157, 0.8627451 , 0.94117647, 0.31372549, 0.58823529,\n",
              "        1.        , 0.89803922, 0.86666667, 0.7372549 , 0.60392157,\n",
              "        0.74901961, 0.82352941, 0.8       , 0.81960784, 0.87058824,\n",
              "        0.89411765, 0.88235294, 0.        ],\n",
              "       [0.38431373, 0.91372549, 0.77647059, 0.82352941, 0.87058824,\n",
              "        0.89803922, 0.89803922, 0.91764706, 0.97647059, 0.8627451 ,\n",
              "        0.76078431, 0.84313725, 0.85098039, 0.94509804, 0.25490196,\n",
              "        0.28627451, 0.41568627, 0.45882353, 0.65882353, 0.85882353,\n",
              "        0.86666667, 0.84313725, 0.85098039, 0.8745098 , 0.8745098 ,\n",
              "        0.87843137, 0.89803922, 0.11372549],\n",
              "       [0.29411765, 0.8       , 0.83137255, 0.8       , 0.75686275,\n",
              "        0.80392157, 0.82745098, 0.88235294, 0.84705882, 0.7254902 ,\n",
              "        0.77254902, 0.80784314, 0.77647059, 0.83529412, 0.94117647,\n",
              "        0.76470588, 0.89019608, 0.96078431, 0.9372549 , 0.8745098 ,\n",
              "        0.85490196, 0.83137255, 0.81960784, 0.87058824, 0.8627451 ,\n",
              "        0.86666667, 0.90196078, 0.2627451 ],\n",
              "       [0.18823529, 0.79607843, 0.71764706, 0.76078431, 0.83529412,\n",
              "        0.77254902, 0.7254902 , 0.74509804, 0.76078431, 0.75294118,\n",
              "        0.79215686, 0.83921569, 0.85882353, 0.86666667, 0.8627451 ,\n",
              "        0.9254902 , 0.88235294, 0.84705882, 0.78039216, 0.80784314,\n",
              "        0.72941176, 0.70980392, 0.69411765, 0.6745098 , 0.70980392,\n",
              "        0.80392157, 0.80784314, 0.45098039],\n",
              "       [0.        , 0.47843137, 0.85882353, 0.75686275, 0.70196078,\n",
              "        0.67058824, 0.71764706, 0.76862745, 0.8       , 0.82352941,\n",
              "        0.83529412, 0.81176471, 0.82745098, 0.82352941, 0.78431373,\n",
              "        0.76862745, 0.76078431, 0.74901961, 0.76470588, 0.74901961,\n",
              "        0.77647059, 0.75294118, 0.69019608, 0.61176471, 0.65490196,\n",
              "        0.69411765, 0.82352941, 0.36078431],\n",
              "       [0.        , 0.        , 0.29019608, 0.74117647, 0.83137255,\n",
              "        0.74901961, 0.68627451, 0.6745098 , 0.68627451, 0.70980392,\n",
              "        0.7254902 , 0.7372549 , 0.74117647, 0.7372549 , 0.75686275,\n",
              "        0.77647059, 0.8       , 0.81960784, 0.82352941, 0.82352941,\n",
              "        0.82745098, 0.7372549 , 0.7372549 , 0.76078431, 0.75294118,\n",
              "        0.84705882, 0.66666667, 0.        ],\n",
              "       [0.00784314, 0.        , 0.        , 0.        , 0.25882353,\n",
              "        0.78431373, 0.87058824, 0.92941176, 0.9372549 , 0.94901961,\n",
              "        0.96470588, 0.95294118, 0.95686275, 0.86666667, 0.8627451 ,\n",
              "        0.75686275, 0.74901961, 0.70196078, 0.71372549, 0.71372549,\n",
              "        0.70980392, 0.69019608, 0.65098039, 0.65882353, 0.38823529,\n",
              "        0.22745098, 0.        , 0.        ],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.15686275, 0.23921569, 0.17254902,\n",
              "        0.28235294, 0.16078431, 0.1372549 , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        ],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        ],\n",
              "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
              "        0.        , 0.        , 0.        ]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9-Two7wicJ8N",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WsZjP3bacaUp",
        "colab_type": "text"
      },
      "source": [
        "Reshaping the images:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D7RYTbvWccwX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_images = train_images.reshape((60000,28,28,1))\n",
        "test_images = test_images.reshape((10000,28,28,1))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "k29A0n1yerMI",
        "colab_type": "text"
      },
      "source": [
        "Importing appropriate tools for modeling our neural network:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HLGrAv-oe4fp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Conv2D, Flatten, Dense, MaxPooling2D, Dropout"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Sr_HR67bgNWU",
        "colab_type": "text"
      },
      "source": [
        "Modeling a convulotional neural network:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fPkKGJFNhYIj",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 386
        },
        "outputId": "cf45af08-0efe-4fbd-f3ba-cb4295b17d40"
      },
      "source": [
        "model = Sequential()\n",
        "model.add(Conv2D(32,(5,5),padding='same',input_shape=(28,28,1),activation='relu'))\n",
        "model.add(MaxPooling2D(pool_size=(2,2)))\n",
        "model.add(Conv2D(64,(5,5),padding='same',activation='relu'))\n",
        "model.add(MaxPooling2D(pool_size=(2,2)))\n",
        "model.add(Flatten())\n",
        "model.add(Dense(128,activation='relu'))\n",
        "model.add(Dense(10,activation='softmax'))\n",
        "model.summary()"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_3\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_4 (Conv2D)            (None, 28, 28, 32)        832       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_4 (MaxPooling2 (None, 14, 14, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_5 (Conv2D)            (None, 14, 14, 64)        51264     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_5 (MaxPooling2 (None, 7, 7, 64)          0         \n",
            "_________________________________________________________________\n",
            "flatten_1 (Flatten)          (None, 3136)              0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 128)               401536    \n",
            "_________________________________________________________________\n",
            "dense_3 (Dense)              (None, 10)                1290      \n",
            "=================================================================\n",
            "Total params: 454,922\n",
            "Trainable params: 454,922\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JO03X6eLpIqg",
        "colab_type": "text"
      },
      "source": [
        "Compile the model:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eCHdTPI5pK0j",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.compile(\n",
        "    optimizer='adam',\n",
        "    loss='sparse_categorical_crossentropy',\n",
        "    metrics=['accuracy'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Jmflq74Qp_ao",
        "colab_type": "text"
      },
      "source": [
        "Early stopping callback:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TZn1kVzaqJfx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.callbacks import EarlyStopping"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "E6-Mplp9qVk7",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "early_stop = EarlyStopping(monitor='val_loss',patience=2)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dV6qBissq-n9",
        "colab_type": "text"
      },
      "source": [
        "Training the model by fitting it to the train data:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u-tSNwzNq_JX",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 689
        },
        "outputId": "268f6eb2-e535-45ca-c081-f4d7e3eeb47a"
      },
      "source": [
        "history = model.fit(train_images,train_labels,epochs=20,validation_split=0.2)"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/20\n",
            "1500/1500 [==============================] - 11s 7ms/step - loss: 0.4304 - accuracy: 0.8448 - val_loss: 0.3291 - val_accuracy: 0.8817\n",
            "Epoch 2/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.2820 - accuracy: 0.8968 - val_loss: 0.2797 - val_accuracy: 0.9041\n",
            "Epoch 3/20\n",
            "1500/1500 [==============================] - 11s 7ms/step - loss: 0.2377 - accuracy: 0.9134 - val_loss: 0.2563 - val_accuracy: 0.9091\n",
            "Epoch 4/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.2046 - accuracy: 0.9232 - val_loss: 0.2366 - val_accuracy: 0.9145\n",
            "Epoch 5/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.1769 - accuracy: 0.9336 - val_loss: 0.2295 - val_accuracy: 0.9182\n",
            "Epoch 6/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.1537 - accuracy: 0.9427 - val_loss: 0.2490 - val_accuracy: 0.9148\n",
            "Epoch 7/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.1325 - accuracy: 0.9496 - val_loss: 0.2533 - val_accuracy: 0.9178\n",
            "Epoch 8/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.1107 - accuracy: 0.9587 - val_loss: 0.2668 - val_accuracy: 0.9132\n",
            "Epoch 9/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0983 - accuracy: 0.9630 - val_loss: 0.2787 - val_accuracy: 0.9156\n",
            "Epoch 10/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0871 - accuracy: 0.9679 - val_loss: 0.2937 - val_accuracy: 0.9193\n",
            "Epoch 11/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0712 - accuracy: 0.9736 - val_loss: 0.3091 - val_accuracy: 0.9178\n",
            "Epoch 12/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0625 - accuracy: 0.9755 - val_loss: 0.3754 - val_accuracy: 0.9041\n",
            "Epoch 13/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0547 - accuracy: 0.9798 - val_loss: 0.3727 - val_accuracy: 0.9114\n",
            "Epoch 14/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0508 - accuracy: 0.9816 - val_loss: 0.4112 - val_accuracy: 0.9165\n",
            "Epoch 15/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0426 - accuracy: 0.9843 - val_loss: 0.4181 - val_accuracy: 0.9148\n",
            "Epoch 16/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0415 - accuracy: 0.9844 - val_loss: 0.4701 - val_accuracy: 0.9171\n",
            "Epoch 17/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0383 - accuracy: 0.9866 - val_loss: 0.4574 - val_accuracy: 0.9183\n",
            "Epoch 18/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0314 - accuracy: 0.9883 - val_loss: 0.4650 - val_accuracy: 0.9099\n",
            "Epoch 19/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0347 - accuracy: 0.9868 - val_loss: 0.5031 - val_accuracy: 0.9158\n",
            "Epoch 20/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0307 - accuracy: 0.9891 - val_loss: 0.5183 - val_accuracy: 0.9154\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YhhTLC8ABe7E",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 151
        },
        "outputId": "1c928e68-1c03-492f-87d6-dacee02ccbaf"
      },
      "source": [
        "history = model.fit(train_images,train_labels,epochs=20,validation_split=0.2,callbacks=early_stop)"
      ],
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0243 - accuracy: 0.9913 - val_loss: 0.6202 - val_accuracy: 0.9119\n",
            "Epoch 2/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0252 - accuracy: 0.9908 - val_loss: 0.5984 - val_accuracy: 0.9142\n",
            "Epoch 3/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0273 - accuracy: 0.9908 - val_loss: 0.6126 - val_accuracy: 0.9144\n",
            "Epoch 4/20\n",
            "1500/1500 [==============================] - 10s 7ms/step - loss: 0.0191 - accuracy: 0.9939 - val_loss: 0.6761 - val_accuracy: 0.9137\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kJRtsQlhxLXq",
        "colab_type": "text"
      },
      "source": [
        "Evaluate the model with the test data:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "adXNuDg3xLvf",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "3cf69f06-2627-4c5c-b902-aa0d11565174"
      },
      "source": [
        "test_loss,test_accuracy = model.evaluate(test_images,test_labels)"
      ],
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "313/313 [==============================] - 1s 4ms/step - loss: 0.6784 - accuracy: 0.9149\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zCWxi9qLyeV_",
        "colab_type": "text"
      },
      "source": [
        "Loss plot:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "34sXlvi2ygIY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title(\"Loss\")\n",
        "plt.ylabel(\"loss\")\n",
        "plt.xlabel(\"epoch\")\n",
        "plt.legend(['train', 'test'], loc='upper right')\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3m9HRlcazeYE",
        "colab_type": "text"
      },
      "source": [
        "Accuracy plot:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GtolTw3Jzgnj",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "outputId": "f701c79d-7bf9-438e-d8c5-d90571bbb21d"
      },
      "source": [
        "plt.plot(history.history['accuracy'])\n",
        "plt.plot(history.history['val_accuracy'])\n",
        "plt.title(\"Accuracy\")\n",
        "plt.ylabel(\"accuracy\")\n",
        "plt.xlabel(\"epoch\")\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()"
      ],
      "execution_count": 35,
      "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": {
        "id": "PoTZjOQiDKaL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}