Tensorflow_MNIST.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Tensorflow_MNIST.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Damntoochill/Learning-ML/blob/master/Tensorflow_MNIST.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "metadata": {
        "id": "1hvTKY1cpSRC",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#importing dependencies\n",
        "import tensorflow as tf\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "hAEdp5Fop-Zd",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#downloading data directly\n",
        "data = tf.keras.datasets.mnist\n",
        "\n",
        "#dividing the data into train and test\n",
        "(x_train,y_train), (x_test,y_test) = data.load_data()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "9GAOrbQdqbbh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 347
        },
        "outputId": "8c7f8dee-edd8-4881-b147-61b2169a0a05"
      },
      "cell_type": "code",
      "source": [
        "plt.imshow(x_train[0])\n",
        "plt.show()"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUsAAAFKCAYAAACU6307AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "ixcr9b_1q3xq",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1046
        },
        "outputId": "6e0db9ea-bd78-4cad-db5d-19faddc8e43d"
      },
      "cell_type": "code",
      "source": [
        "#the dataset is nothing but an array of values. So lets see what's in it\n",
        "print(x_train[0])"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   3  18  18  18 126 136\n",
            "  175  26 166 255 247 127   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0  30  36  94 154 170 253 253 253 253 253\n",
            "  225 172 253 242 195  64   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0  49 238 253 253 253 253 253 253 253 253 251\n",
            "   93  82  82  56  39   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0  18 219 253 253 253 253 253 198 182 247 241\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0  80 156 107 253 253 205  11   0  43 154\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0  14   1 154 253  90   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0 139 253 190   2   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0  11 190 253  70   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0  35 241 225 160 108   1\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0  81 240 253 253 119\n",
            "   25   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0  45 186 253 253\n",
            "  150  27   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0  16  93 252\n",
            "  253 187   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0 249\n",
            "  253 249  64   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0  46 130 183 253\n",
            "  253 207   2   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0  39 148 229 253 253 253\n",
            "  250 182   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0  24 114 221 253 253 253 253 201\n",
            "   78   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0  23  66 213 253 253 253 253 198  81   2\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0  18 171 219 253 253 253 253 195  80   9   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0  55 172 226 253 253 253 253 244 133  11   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 136 253 253 253 212 135 132  16   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "5Wu8j3dWsFOF",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "x_train[0] is a 28x28 matrix. The values in the matrix range from 0 - 255. The neural network can learn faster if the values are between (0,,1). So, let's normalise it."
      ]
    },
    {
      "metadata": {
        "id": "FxNxjRCQsB2K",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#normalization\n",
        "x_train = tf.keras.utils.normalize(x_train, axis = 0)\n",
        "x_test = tf.keras.utils.normalize(x_test,axis = 0)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "0ZMBlyG4yMeT",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 347
        },
        "outputId": "49cbadc5-8932-4395-8f4b-f7d903352ebd"
      },
      "cell_type": "code",
      "source": [
        "#let's check how the image looks after normalization\n",
        "plt.imshow(x_train[0])\n",
        "plt.show()"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUsAAAFKCAYAAACU6307AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nDP4PdDcZHAIXwUYAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "XHTfQtQRyZpy",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "why do some pixels are still black? Let's check"
      ]
    },
    {
      "metadata": {
        "id": "Z_8-7SHKyuLp",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 3617
        },
        "outputId": "26dfdf3d-8a1d-4356-fd96-53ba82d1c150"
      },
      "cell_type": "code",
      "source": [
        "print(x_train[0])"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  9.38754405e-05 5.19970263e-04 5.53561164e-04 5.37029201e-04\n",
            "  3.47166061e-03 3.51428135e-03 5.94251333e-03 1.23168058e-03\n",
            "  1.19785099e-02 1.79321461e-02 2.20310048e-02 3.04464368e-02\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  1.69714608e-03 1.73366984e-03 3.83996450e-03 4.82944767e-03\n",
            "  4.30112118e-03 5.97840981e-03 6.49906235e-03 6.39548152e-03\n",
            "  5.93555243e-03 5.49638318e-03 6.29811101e-03 6.63932407e-03\n",
            "  1.47119406e-02 1.36565700e-02 1.41051354e-02 1.22143395e-02\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 3.20546788e-03\n",
            "  1.09503608e-02 1.00306160e-02 8.65621518e-03 6.76150056e-03\n",
            "  5.59973186e-03 5.41955652e-03 6.10575819e-03 6.13500282e-03\n",
            "  5.72073197e-03 5.19740552e-03 2.41488990e-03 2.84010189e-03\n",
            "  4.21558562e-03 2.80530819e-03 2.51895708e-03 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 1.03101767e-03\n",
            "  8.75369259e-03 8.70035206e-03 7.57318862e-03 6.08622194e-03\n",
            "  5.28933716e-03 5.38393925e-03 4.95918542e-03 4.65265368e-03\n",
            "  5.84204856e-03 5.12032105e-03 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  3.01359988e-03 5.04781304e-03 3.05949494e-03 6.05573469e-03\n",
            "  5.53888938e-03 4.80224236e-03 3.13255131e-04 0.00000000e+00\n",
            "  1.11235328e-03 3.48821193e-03 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 4.52075889e-04 2.93432040e-05 3.96492381e-03\n",
            "  6.26720143e-03 2.46981965e-03 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 3.89714459e-03\n",
            "  7.02274593e-03 5.76715856e-03 6.85173816e-05 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 3.19507815e-04\n",
            "  5.39920551e-03 7.38671425e-03 2.14523166e-03 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  9.30897194e-04 6.11876850e-03 5.93924156e-03 4.03969385e-03\n",
            "  2.50585452e-03 2.13776301e-05 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 1.78965971e-03 5.66253254e-03 5.90080421e-03\n",
            "  5.63419877e-03 2.53081374e-03 7.20308629e-04 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 1.01851575e-03 4.29732080e-03\n",
            "  5.72469346e-03 5.59657282e-03 4.51456815e-03 1.13917202e-03\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 3.86269330e-04\n",
            "  2.21561716e-03 5.97324381e-03 7.91793923e-03 7.89679429e-03\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 6.20106903e-03 8.02984946e-03 1.04157723e-02\n",
            "  3.94961875e-03 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 1.26341173e-03 3.51235725e-03\n",
            "  4.75456404e-03 6.26888840e-03 7.90325134e-03 8.66569312e-03\n",
            "  1.26285991e-04 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  9.87370681e-04 3.64955319e-03 6.20887767e-03 6.74057329e-03\n",
            "  6.41335965e-03 6.09051247e-03 7.81078925e-03 7.88346563e-03\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 7.45647690e-04 3.03527130e-03\n",
            "  5.15639333e-03 5.80420628e-03 6.49136867e-03 6.49078848e-03\n",
            "  6.25676434e-03 4.86359736e-03 2.54136205e-03 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  7.44782099e-04 1.98694467e-03 6.09764352e-03 6.22090475e-03\n",
            "  5.51742209e-03 5.51681356e-03 6.30109419e-03 5.07703688e-03\n",
            "  2.06162879e-03 5.18537791e-05 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 1.24441601e-03 8.29596104e-03\n",
            "  7.31808257e-03 7.58140648e-03 7.08009467e-03 6.08286502e-03\n",
            "  5.46858656e-03 4.29589459e-03 2.06534267e-03 2.48910286e-04\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  2.48800395e-02 2.46111047e-02 2.07362668e-02 1.55490554e-02\n",
            "  1.02011297e-02 8.84654678e-03 8.08588535e-03 6.62369919e-03\n",
            "  3.22177712e-03 2.73407964e-04 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  9.94299723e-02 5.61903233e-02 3.53237440e-02 2.34054087e-02\n",
            "  1.29693369e-02 7.08436987e-03 6.27602291e-03 6.32989798e-04\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]\n",
            " [0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00\n",
            "  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "0TeMHMfmzD8k",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "Can't understand why. Let's not waste time on this :("
      ]
    },
    {
      "metadata": {
        "id": "MyxxjvEHxWRb",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#creating a model\n",
        "model = tf.keras.models.Sequential()\n",
        "model.add(tf.keras.layers.Flatten())#Flattens the input\n",
        "model.add(tf.keras.layers.Dense(128, activation = tf.nn.relu))#hidden layer 1\n",
        "model.add(tf.keras.layers.Dense(128, activation = tf.nn.relu))#hidden layer 2\n",
        "model.add(tf.keras.layers.Dense(10, activation = tf.nn.softmax))#output layer"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "oIcz5CxH0wbm",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "**Flatten** : If inputs are shaped (batch,) without a channel dimension, then flattening adds an extra channel dimension and output shapes are (batch, 1)."
      ]
    },
    {
      "metadata": {
        "id": "_yXHymeO02ZI",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 92
        },
        "outputId": "bf4bcea0-8653-40b6-db04-9488b249057d"
      },
      "cell_type": "code",
      "source": [
        "#compiling the model\n",
        "model.compile(optimizer = 'adam', loss = \"sparse_categorical_crossentropy\", metrics = ['accuracy'])"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/resource_variable_ops.py:435: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "LhGffrDH3lDE",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 146
        },
        "outputId": "0e5fa46e-0306-4524-cc04-ac75ada1454d"
      },
      "cell_type": "code",
      "source": [
        "#training the model\n",
        "model.fit(x_train, y_train,epochs = 3)"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/3\n",
            "60000/60000 [==============================] - 9s 155us/sample - loss: 0.5686 - acc: 0.8334\n",
            "Epoch 2/3\n",
            "60000/60000 [==============================] - 9s 147us/sample - loss: 0.2677 - acc: 0.9216\n",
            "Epoch 3/3\n",
            "60000/60000 [==============================] - 10s 162us/sample - loss: 0.2080 - acc: 0.9388\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7ff9589b14e0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "metadata": {
        "id": "4OYQFbOA39BO",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        },
        "outputId": "c425cc3b-e22f-4744-ea08-7e53d817064d"
      },
      "cell_type": "code",
      "source": [
        "#evaluating the model on the test set\n",
        "val_loss, val_acc = model.evaluate(x_test,y_test)\n",
        "print(val_loss, val_acc)"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "10000/10000 [==============================] - 1s 53us/sample - loss: 0.2659 - acc: 0.9398\n",
            "0.26588178882119246 0.9398\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "MPlnA8r647ZI",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#saving the model\n",
        "model.save('mnist.model')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "5PGSTJXw54HP",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 55
        },
        "outputId": "6487deee-2ea8-463f-8a86-0a4c6d29b784"
      },
      "cell_type": "code",
      "source": [
        "#loading the saved model\n",
        "new_model = tf.keras.models.load_model('mnist.model')"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:Sequential models without an `input_shape` passed to the first layer cannot reload their optimizer state. As a result, your model isstarting with a freshly initialized optimizer.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "o2080j2-6j9x",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "x_test = np.array(x_test)\n",
        "new_Predictions = new_model.predict(x_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "AQ6QVBrQ7qMn",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 256
        },
        "outputId": "4c8c6c95-b391-472d-ddc7-a6f25bc33c9e"
      },
      "cell_type": "code",
      "source": [
        "print(new_Predictions)"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[4.06095673e-08 1.46321909e-14 2.78161260e-05 ... 9.99801815e-01\n",
            "  2.81718826e-09 1.85971061e-07]\n",
            " [1.26265098e-09 4.34493970e-08 9.99895811e-01 ... 3.12376590e-18\n",
            "  1.89467659e-08 4.97102926e-19]\n",
            " [1.99748662e-10 9.99675989e-01 2.17600507e-04 ... 3.80896518e-06\n",
            "  8.07709876e-05 5.21364640e-08]\n",
            " ...\n",
            " [2.54005428e-14 3.61750662e-14 2.97960137e-12 ... 4.25281357e-08\n",
            "  1.04135936e-04 7.66355661e-04]\n",
            " [5.93329330e-10 1.29101313e-10 4.56687029e-14 ... 8.33674560e-14\n",
            "  3.76357697e-03 4.92300256e-10]\n",
            " [2.35329534e-10 7.29930849e-19 9.55784757e-11 ... 4.61628830e-22\n",
            "  2.60832039e-14 3.79957724e-20]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "1Bb91q818CMM",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "837353bf-6b19-4b77-fb4d-be10a54d68a2"
      },
      "cell_type": "code",
      "source": [
        "print(np.argmax(new_Predictions[2]))"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "Q0XU09fV8Pa0",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "71b3134a-3e43-468e-fd68-2c5dc06d36ca"
      },
      "cell_type": "code",
      "source": [
        "print(y_test[2])"
      ],
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "cYEXj0Qb8Wyv",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}