MSDS - Computer Vision/Assignment_1_Fashion_MNIST.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Assignment 1 - Fashion MNIST",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyMKD7gJJvM3776OTUByKZaz",
      "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/tiffduong/Data_Science_Portfolio/blob/master/MSDS%20-%20Computer%20Vision/Assignment_1_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": "ky0v3Qqrqhex",
        "colab_type": "text"
      },
      "source": [
        "# **Assignment 1 - Classification Model with Fashion MNIST data**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HKfR7X33KRFm",
        "colab_type": "text"
      },
      "source": [
        "# Data Set-Up and Preprocessing"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZQbt2E4GqPJa",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# import dependencies\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from keras.utils import to_categorical"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZShdgoderZMi",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# download fashion mnist data \n",
        "(X_train, y_train), (X_test, y_test) = keras.datasets.fashion_mnist.load_data()\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j8DLZ8OQGcRk",
        "colab_type": "code",
        "outputId": "4902fc5a-40f4-4a95-8786-80566eaf9d81",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "print(\"Train Set Shape: {}\".format(X_train.shape))\n",
        "print(\"Test Set Shape: {}\".format(X_test.shape))"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train Set Shape: (60000, 28, 28)\n",
            "Test Set Shape: (10000, 28, 28)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Lt1_n5g3GwmG",
        "colab_type": "code",
        "outputId": "4ed0fb45-d1c5-427a-d1ca-d80af35e14b2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 593
        }
      },
      "source": [
        "# plot first few images from the training set\n",
        "plt.figure(figsize=(10,10))\n",
        "for i in range(9):\n",
        "\t# define subplot\n",
        "\tplt.subplot(330 + 1 + i)\n",
        "\t# plot raw pixel data\n",
        "\tplt.imshow(X_train[i], cmap=plt.get_cmap('tab20b'))\n",
        "# show the figure\n",
        "plt.show()\n"
      ],
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAksAAAJACAYAAACdeiLBAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nOzdf3Dd1X3n/9e72GCMASkLgRtMF3+p\nm5EGGssYpnaajKuUDuvWS2IPTImG8kMbMxu7BJrMN3xTL6SpW9gdE3aITQZTuQZWoQNrZ4O/8bR1\nozCmWMnYlry1I4UAMWnsvcZhYgEGlNjJ+f7hy3d8+bzPx/ej+0P3fvR8zDCW3vfofM69um/pcHXe\n920hBAEAAMD3G5O9AAAAgGbGZgkAACAFmyUAAIAUbJYAAABSsFkCAABIwWYJAAAgRVWbJTO7zsxe\nNLOXzeyeWi0KaFXkBFCOnEAe2ETfZ8nMzpD0I0nXSjooaZekm0III7GvmT59ZjjrrPMndD2gWm+/\nffj1EMKF9ZqfnPBdcMEFbnzWsdfc+BtnHEvExn/D//+6c86aU/E6ps/0rxcO/cKf+yL/mv/2ensi\ndvyX71a8jmZCTtTX9DPPduMfPO+IGz/y5gfdeKOfX7GcjeVQ8d/OqOdyGiqWE9OqmPMaSS+HEH4s\nSWb295KulxRNgrPOOl+/M++2Ki4JTNzgC/f/pM6XICccvb2fceOLnn/QjW+btTMRG501wx27cG5/\nxesodPnXG//SATe+6G7/mis3fCoRKxaj3+KmRk7UV6HQ6cZXLnnEja/f9lk33ujnVyxnYzm0ZtV5\n9VxOQ8Vyopo/w10i6aenfH6wFAOmKnICKEdOIBfqfsDbzFaY2W4z2338xDv1vhzQ9MgJoBw5gWZX\nzWbpkKRLT/l8dilWJoSwIYSwIISwYPq0mVVcDmh65ARQjpxALlSzWdolaa6ZzTGzMyX9iaRna7Ms\noCWRE0A5cgK5MOED3iGEE2a2StI/SjpD0sYQwg9qtjKgxeQxJ2IHPf9o9hcSsYGvd7tjR55c78bP\nbutw492zliViy+YWIiscj8Qdwyvd8OCMATe+7n7/4Pfam69KxMaGP+SO3Ta82o1ffHvyYHpf32Pu\n2FaWx5zIYvX9fqXl5hWL3HiH9rrxNb9d+fN89LD/vO24OFk52n51lzt2YOs2Nz7/Sj9nV68bTcTy\ndOhbqq4aTiGEbZL8RxWYgsgJoBw5gTzgHbwBAABSsFkCAABIwWYJAAAgRVVnlgC0Hu9dhXvefsod\nOx456Dmg5GHukTH/YOmqG/xDoXHOPOP+3OueSR4sjYmto/ta/526uxVZ93jyHcb79vvr6Lywx5/D\neVz/yJ5zh347LPbnQNMrHht24+23Fd14z7Elbnxge/KA98KlfkGFf2RbGhtOXnNgu/+8PfAz/53x\n+0f9QouOK2MFGPnBK0sAAAAp2CwBAACkYLMEAACQgs0SAABACjZLAAAAKaiGywGvumnRf/ymO3bz\nox+u93LQ5Nac+71EbPhXfhVOrFLMHRurHquj7JV29VGLdQxs9yvnepf635s8tkfJm8IsvzZtVH41\nXPFaf55uJfNwYKvfpifGq1Y9+3iyslOSDnzMn2Plwki13qBXNZuvdie8sgQAAJCCzRIAAEAKNksA\nAAAp2CwBAACkYLMEAACQgmq4FjL/qhvd+Oo//qtEbOSgX20x+nE/HjOyY3Gm8WgenR9/zo0P/zBZ\ndZWl6i0m1qdt+RV+BU1bV2v2k/J6bG3e7/fQq0WV3B/N/oIb7xOVrc0u1hsu1ktt/ca+iudedu98\nNz660a9wa//UnOTYSGvFjo7WzM164pUlAACAFGyWAAAAUrBZAgAASMFmCQAAIEVVB7zN7FVJb0n6\nlaQTIYQFtVgU0KrICaAcOYE8qEU13O+HEF6vwTw4jTnhdje+7v7uRGz+jHF37Jq7I1VP7ck5JGlk\nbn8itn7bZ92xxeKIP/fU0xQ58em5fvzwD5OxWCVbTdyZ7EklSVsG17vxthNrE7GxaX5FmDc2q8GX\n/J5sMR1K5lZnW2+mObzHu7MtWa0kSQNf93NTOpTpmpOsKXKi0Ub3+dXHsWq4ZTf6FW5Z5h6d5f+M\n9+oys1xvquPPcAAAACmq3SwFSf9kZnvMbEUtFgS0OHICKEdOoOVV+2e43wshHDKzD0rabmY/DCHs\nOHVAKTlWSNKZZ51X5eWApkdOAOXICbS8ql5ZCiEcKv17RNI3JV3jjNkQQlgQQlgwfdrMai4HND1y\nAihHTiAPJvzKkpmdI+k3QghvlT7+Q0lfqdnKkDA+7B/cW96VbCexeX9kkof8g7wHfst/m33vmvP1\niDt2xnz/ULmuTh5c3fnsp9yhvX/xPTc+/2f+Ycb+w8lYbO56a7ac6Bz3Dx0fdg4ox8TadXgHlGNj\nl936ihtffod/zfkfG0jEVt7qt/ZYPn+5G/cOucZaOMTiscPjfX2PJWKrPugOzXRwfmTMPwjfe7P/\nfdy59xdufGjP0xVfs96aLScaLXaQu55z1/OaU1k1f4a7SNI3zey9eb4RQviHmqwKaE3kBFCOnEAu\nTHizFEL4saSP1HAtQEsjJ4By5ATygrcOAAAASMFmCQAAIAWbJQAAgBS1aHeCGlt+x4tu/OjfLXLj\nowcrb50wNBa56H6/kuls7UzEFvpdVyRFWqmcSFay9XzxjchYfx1DF/rD5zzkVBDd9E137Jb/6c+R\nV+/O8J8vUrLaLCZLNVd87JludNmJNf7w55OhLfdtc4e+ezDZjkeShp9Jtkfpmhu5XkTfk347Fu/+\nxCrZYnkYG+8Z3Op/v3r/m59vQ7dWPDUmSWFWlxsvHhtu8EpqYypU4PHKEgAAQAo2SwAAACnYLAEA\nAKRgswQAAJCCzRIAAEAKquGaUM+xZK83SSpe4VfW1FfymgNPVl7JI0lnH09W1Ol2v4opZn6kz9nA\njGT13OZHD2Wau9X19n7GjQ9u9SvIslRixXhVXlnnHYhUTw7vTVZKjr/k9wY8+tIef/JzbkqE1vY9\n6Q4tFDrd+KLbIvfn0WSfup7L/H57/a9W/1jHHtcDtzl5JUlaXPU1URuxqje0Hl5ZAgAASMFmCQAA\nIAWbJQAAgBRslgAAAFJwwLsJHd3lv+X96Jh/iNQ7ABprsxDTMdtvM5KllUpccvzhrRlnWOq37pgx\nnmxrIX0k2+Qt7vDGHje+pMtv75HlIHaWdh2rbvCfQ4d/2ufGx6b5bX2W3JZsnVCY5R/wLx77nhv3\njO7zD4lL/jq6F/qFFh1XJvOz/88jrX4ishyQj38PIpP/MtNSMAlata3JVMYrSwAAACnYLAEAAKRg\nswQAAJCCzRIAAECK026WzGyjmR0xs/2nxD5gZtvN7KXSv+31XSbQPMgJoBw5gbyrpBpuk6R1kp44\nJXaPpO+EEB4ws3tKn3+x9submtqvjrxF/na/Gi5LZU2sYunori1ufGQsW5VPpbJX1PkWLV2WDPa/\nUpO5U2xSE+XEjC7/eTHjmP897WxLPgdi1ZBtXcnKNEkaedKvcPP8xdzI2BORL8hQKNQmv0rS49/D\nFM9H4sPJ5+6qG/x8W/dMspo0q+5r/Rwsjg74XzB0XdXXnIBNaqKcQH3EqvhGN3qtdybleVg3p31l\nKYSwQ9LP3xe+XtLjpY8fl/TJGq8LaFrkBFCOnEDeTfTM0kUhhPfetOSwpItqtB6gVZETQDlyArlR\n9QHvEEKQFGK3m9kKM9ttZruPn3in2ssBTY+cAMqRE2h1E90svWZmBUkq/XskNjCEsCGEsCCEsGD6\ntJkTvBzQ9MgJoBw5gdyY6GbpWUm3lD6+RdK3arMcoGWRE0A5cgK5cdpqODN7StJiSReY2UFJ90l6\nQNLTZtYr6SeSbqznIqeagUjVW0wtesONHo71pUr2x8raw8qTdY6zx71qC7/aqPPj/tjBFypc3Gk0\nW05sjlQ/bY6MLxTOT8TaZ612x87Z6M/ReaHfj87TF6mci1VmNruznWXHqt7OPh553h4504l6MWkg\nWt3ZPNVGzZYTk6FQ6HSiv2j4OiZDV1eyYnPz0CQspI5Ou1kKIdwUuekTNV4L0BLICaAcOYG84x28\nAQAAUrBZAgAASMFmCQAAIAWbJQAAgBSV9IZDHa3fdHki1va83+9qcKvfC6oWfdbcHmvKVpmXpcIt\ntuaFS7v9udXrxufPSPbm3BZbICRJxeKIE1vsjl3Y+xl/kq31e5Rr0U+tnrwqvtjzuTg71r+tlitC\nM/DySkr+fM+j4eFsFdytiFeWAAAAUrBZAgAASMFmCQAAIAWbJQAAgBQc8J5k6zcmW0GsPJbtwHbH\n7OSB09GD/iHZ4mi2uUfGkgd5sx4oz7K+2CH2kTH/0PvyDcm5R26NtYcYjMSRlXeYv1ut2b6knrrv\nXOnGN0efowCaEa8sAQAApGCzBAAAkILNEgAAQAo2SwAAACnYLAEAAKSgGm6SrTy2JhGLVYrFZBmf\ndW6v8i1LW5PYNbPO0Xuz3+5k9Nh6J3qeOxbZ9fU95sa39EzdyjevHUu03cmx4cgsPEeRH17VZ94q\nPnllCQAAIAWbJQAAgBRslgAAAFKwWQIAAEhx2s2SmW00syNmtv+U2JfN7JCZ7S39t6S+ywSaBzkB\nlCMnkHeVVMNtkrRO0hPviz8UQlhb8xW1iM6PP+fG1yzrcePr/m+/+itLpVh0LRl6tdVi7qy94epp\n+M+GnOjiel92k6Z4Thzdlazyar+6yx276ga/cs6rKmsmtVj36L5i5JbcVcNt0hTPiUKh04n+ouHr\nQH2c9pWlEMIOST9vwFqAlkBOAOXICeRdNWeWVpnZv5Zefm2v2YqA1kVOAOXICeTCRDdLX5d0uaR5\nkoqSHowNNLMVZrbbzHYfP/HOBC8HND1yAihHTiA3JrRZCiG8FkL4VQjh15Iek3RNytgNIYQFIYQF\n06fNnOg6gaZGTgDlyAnkyYTanZhZIYTw3snFT0nanza+EfzDddKSJR9Nju3y/wdndONOf45jixKx\n4R/6B7kH/sZfX5ZD0bGx3dfOcONZDpwuv8IvSIm1QemYnTzkeuBl//HruNhf94yPJd8Kf2FkfWPD\n/oHYWPzd6cnvjX4ZmbyOmjEngMlETiBPTrtZMrOndLK86AIzOyjpPkmLzWyepCDpVUl31HGNQFMh\nJ4By5ATy7rSbpRDCTU64rw5rAVoCOQGUIyeQd7yDNwAAQAo2SwAAACnYLAEAAKSYUDVcrXmtQz5/\neL47tv9VvyKs81dv+JNv3ZYIjW312xgcHfPn7nej2dqGxHgtFfq+4f+p/90ZfnzVDclYrEIuVvW2\ncGl3ZIVJ4wf9irrxWf7jOrx1IBGLtV2JVQJ6VXmS1HtzbyI20PekOxa14+XhqqsnYSGTwMvZge3j\nk7ASoLEKs/yWRgMPr0+OLfy5O7ZYHKnpmhqFV5YAAABSsFkCAABIwWYJAAAgBZslAACAFGyWAAAA\nUjRFNdyaS52Kph/61SWdbdnmjlVdNVqsyqs4mox3fc2vBDz7RLKqTJLe3ZesZOtsy1adM/58sppB\nkra8NJSILZsbqVTcX7/HOlbFp1gcTSNL78JmElu3Vw0X03FloVbLAVpK4UNXuHGq4QAAAHKIzRIA\nAEAKNksAAAAp2CwBAACkYLMEAACQoqHVcL855yyt33R5Il58OFkR1n2tX1kVq1CJVZvF4p5Y/7G2\nrmRFy6DT70ySuq/1+8tlqQjq2LjMjfdf7VesaZqzltn+49fV5a9v9Jl2Nz7nwp5EbHjMX0asX16W\n70EtzL/qRjc++ML9DV0HEOulJb3S0HWg/vzqrz0NX0c9FY8NT/YSJg2vLAEAAKRgswQAAJCCzRIA\nAECK026WzOxSM/uumY2Y2Q/M7HOl+AfMbLuZvVT61z/0AuQMOQGUIyeQd5Uc8D4h6fMhhCEzO1fS\nHjPbLulWSd8JITxgZvdIukfSF9MmeuvYGxoY3JaIH93vHADeX8HKTlGLg9UxI/uTa+65LNZOJHag\n0+ete90zyetJ0nKtdOOFjuTB6oGD/vrmdCVbo0jS2G8V3Xi0zUgGXsuZWh369g7lFxf8VU3mTlGz\nnGhVWVp+TGUDD0eKMnRdQ9fRAFM+J6ay7juTv5v6vjYJC6mj076yFEIohhCGSh+/JWlU0iWSrpf0\neGnY45I+Wa9FAs2EnADKkRPIu0xnlszsMp186eT7ki4KIbz3csRhSRfVdGVACyAngHLkBPKo4s2S\nmc2StFnSXSGEN0+9LYQQJIXI160ws91mtvvtt35R1WKBZlKLnDh+4p0GrBRoDHICeVXRZsnMputk\nAvSHELaUwq+ZWaF0e0HSEe9rQwgbQggLQggLzjn3rFqsGZh0tcqJ6dNmNmbBQJ2RE8izSqrhTFKf\npNEQwldPuelZSbeUPr5F0rdqvzyg+ZATQDlyAnlXSTXcRyXdLGmfme0txb4k6QFJT5tZr6SfSPJ7\nTJzijDBbbSfWJuI9NyTbdRzd5b+tevvVfrVZrOrNq9qJzX3yTGJS97XZKtyyGNierFqLVYrFK9O8\nyiR/bN+TfW48Vt20eX+yki1WCdjV5q9u9LDTBmXcX1/Hxf5973/Vr3b0Ku26u/zKvhqqWU60qlpU\nmeaJ9zyUpO6l/vN281A9VzMppnxOdM07f7KX0FRij8dQi3aAOe1mKYTwL5IscvMnarscoPmRE0A5\ncgJ5xzt4AwAApGCzBAAAkILNEgAAQAo2SwAAACkqqYarmddff119fY8l4nN6n0vEDo/1uHP4nc3i\nFVo7t25JxGIVV1kr7WohVkXjiffjSs5R6PCrcLp2xXra+bzHNfY4RW1Nrm/0raOZppjv3x0NjScf\nk/aP9UZmGcx0TQDA6RWPJSvMC12xqvPz6ruYOuGVJQAAgBRslgAAAFKwWQIAAEjBZgkAACBFQw94\nx6zuW5yILb8j2QJFko5+t+DGY+0wei5LHuZ2229I7kFkyT9cHDsk3izih9IjJ6VfzTA+MjZ2yN57\nrBYtXRa5ni92f+bPSMZX3npmprmRnVdsQAsUTGVj077gRJc0fB3NYsvTsZ4+ixu5jJrhlSUAAIAU\nbJYAAABSsFkCAABIwWYJAAAgBZslAACAFE1RDefZ/OiH/Xhk/PI7XnTj7ceS1Qjj2/22IR2zY+1E\nkmJNQ9q6/Gq9LAa3+pWA2u5ftfvaZMVarDKtnqJVhp5I5eGch/0qucM/3enG1+5YXPk1AQA1V5jl\ntcCKVcO1Jl5ZAgAASMFmCQAAIAWbJQAAgBRslgAAAFKcdrNkZpea2XfNbMTMfmBmnyvFv2xmh8xs\nb+m/qfu+7phSyAmgHDmBvKukGu6EpM+HEIbM7FxJe8xse+m2h0IIa+u3vMrFq+deqXiO3qV+Hg++\n1JOILZzb7471+wNJbSf8h2l47xuJ2NCRQ7Eluta5Szkv0xy18ZHqp7g19v1aXP3ctdMSOVFPO7du\ncaJ+NanXRy6revadi62PXneZTPmccH8H7Vrvju2+c2XV1yseG656jqxGN/pVyaNKxo8W/zwyy0gN\nV9Q4p90shRCKkoqlj98ys1FJl9R7YUCzIieAcuQE8i7TmSUzu0xSl6Tvl0KrzOxfzWyjmbXXeG1A\n0yMngHLkBPKo4s2Smc3SyfeEvCuE8Kakr0u6XNI8nfw/igcjX7fCzHab2e7jJ96pwZKB5kBOAOXI\nCeRVRZslM5uukwnQH0LYIkkhhNdCCL8KIfxa0mOSrvG+NoSwIYSwIISwYPq0mbVaNzCpyAmgHDmB\nPKukGs4k9UkaDSF89ZT4qX09PiVpf+2XBzQfcgIoR04g7yqphvuopJsl7TOzvaXYlyTdZGbzJAVJ\nr0q6oy4rbKC+vscityxOREZ2xMb6VXkn/6cKOTFlciKm4+I5idj4WDImxavKll9ReRX58iv8uWvR\ni7E4XHTjvTcvcuPrnuxLxDrb/PUNRqqHcmjK54Rn89B1/g2D29xw90I/JwYe9qvqGu5q/3nuV6O3\nZtVbTCXVcP8iyZyb/O82kHPkBFCOnEDe8Q7eAAAAKdgsAQAApGCzBAAAkKKSA94AUKb9puSh6O52\nv21I99EDkVmqP7Q6ctA/nJ1Jlx8+GhnefsV4Ita91D/42n84cgD90QrWhdyKtud61G/3VCgkW4cs\nWfJRd6zXnitmxtuPuPGhPU/7XzBU8dS5wytLAAAAKdgsAQAApGCzBAAAkILNEgAAQAo2SwAAACks\nhNC4i5n9TNJPSp9eIOn1hl18cnAfm8u/DyFcONmLOBU5kUutdB/JicnHfWwubk40dLNUdmGz3SGE\nBZNy8QbhPiKLqfBYch+RxVR4LLmPrYE/wwEAAKRgswQAAJBiMjdLGybx2o3CfUQWU+Gx5D4ii6nw\nWHIfW8CknVkCAABoBfwZDgAAIEXDN0tmdp2ZvWhmL5vZPY2+fr2Y2UYzO2Jm+0+JfcDMtpvZS6V/\n2ydzjdUws0vN7LtmNmJmPzCzz5XiubmPk4WcaE3kRP2QE60pzznR0M2SmZ2hk63G/4OkTkk3mVln\nI9dQR5skXfe+2D2SvhNCmCvpO6XPW9UJSZ8PIXRK+l1JK0vfuzzdx4YjJ1r6+UJO1AE50dLPl9zm\nRKNfWbpG0sshhB+HEH4p6e8lXd/gNdRFCGGHpJ+/L3y9pMdLHz8u6ZMNXVQNhRCKIYSh0sdvSRqV\ndIlydB8nCTnRosiJuiEnWlSec6LRm6VLJP30lM8PlmJ5dVEIoVj6+LCkiyZzMbViZpdJ6pL0feX0\nPjYQOZED5ERNkRM5kLec4IB3g4STZYctX3poZrMkbZZ0VwjhzVNvy8t9RGPk5flCTqBW8vJ8yWNO\nNHqzdEjSpad8PrsUy6vXzKwgSaV/j0zyeqpiZtN1MgH6QwhbSuFc3cdJQE60MHKiLsiJFpbXnGj0\nZmmXpLlmNsfMzpT0J5KebfAaGulZSbeUPr5F0rcmcS1VMTOT1CdpNITw1VNuys19nCTkRIsiJ+qG\nnGhRec6Jhr8ppZktkfTfJZ0haWMI4a8buoA6MbOnJC3Wye7Kr0m6T9L/kvS0pN/UyS7aN4YQ3n+4\nryWY2e9Jel7SPkm/LoW/pJN/j87FfZws5ERrPl/IifohJ1rz+ZLnnOAdvAEAAFJwwBsAACAFmyUA\nAIAUbJYAAABSsFkCAABIwWYJAAAgBZslAACAFGyWAAAAUrBZAgAASMFmCQAAIAWbJQAAgBRslgAA\nAFKwWQIAAEjBZgkAACAFmyUAAIAUbJYAAABSVLVZMrPrzOxFM3vZzO6p1aKAVkVOAOXICeSBhRAm\n9oVmZ0j6kaRrJR2UtEvSTSGEkdjXTJ8+M5x11vkTut5kmnlOeyJ2wQffccf+8t+OuvF3f/0rN/4b\nF0xPxI69kbyeJE0764gb/9C5yTmOHD7LHfvB9hlu/Ofj/txH326971fM228ffj2EcGG95p9KOZEn\nXn5L0jtv+7mcJ+REfU0/82w3fsHFx9x45Ee8xo//IhE75yL/tY7xXx534zNmfiAZfDvyesk5v3bD\nr/ybP3eexHJiWhVzXiPp5RDCjyXJzP5e0vWSoklw1lnn63fm3VbFJSfH/KtuTMR6/2yPO/bAnVvc\n+Ohb/g/eGSsKidjOZz/ljm2f+4gbX/P7yTnW3T/HHbvqhg433j+63o1vHrrOjbeiwRfu/0mdLzFl\nciJPvPyWpKE9Tzd4JY1HTtRXodDpxnv/4ntufM7D/jyjhw8kYovu9v/Hd+Rg0Y13XtmTDO7y59DV\n42542V2H/PE5EsuJav4Md4mkn57y+cFSDJiqyAmgHDmBXKj7AW8zW2Fmu81s9/ET/p+ugKmEnADK\nkRNodtVslg5JuvSUz2eXYmVCCBtCCAtCCAumT5tZxeWApkdOAOXICeRCNZulXZLmmtkcMztT0p9I\nerY2ywJaEjkBlCMnkAsTPuAdQjhhZqsk/aOkMyRtDCH8oGYrq6Pld7zoxpd8N3lQWpJGX/qLRGzG\n3/gHqGOWzZ3vxsd3LUuOXeKvY3Crc0BP0sAPk7FVN/gH947uGnbjS7TIH3/mL5Nz/zf/vk+Fw39p\nWjknsur77TcTsf5X/edcz2X+YdH15/4nN16Lg9Xeoe1FP33BHxyJr+653I0v639lwuuaalo5J2KH\ns9ev+Kb/BS8vSYTWPbPXHTp/sMuf42o/PGc0+Xvi3YP+74nYb6Z39yVjg9sH3LHd8nO52/l9IPk5\nXvzYkDt2/bbPuvFiMXrmvylUUw2nEMI2SdtqtBag5ZETQDlyAnnAO3gDAACkYLMEAACQgs0SAABA\niqrOLLUC76Dn+Ab/gFn7vf4ci9qTB/diFsl/h+yYnfc+mAy+7I+NvNeqq+8b/juG9z7kHxKPWXU0\neQCw7+6d7tgt9/kHDpf95YczXRPNr/3q5AHVzjH/IPf4LD8nPj/Pb8kzFv4gESv86ag7tvhEJN/C\nz5Ox2f7Y0YP+3Oue8ePSmZE4WpX3e6Jt3/9wxx59yv85JyWLZ+bPSL7ztiTt3Bp7bvkWfSX5O6jv\n7tWZ5vB+9h/Y2O+OPbrLL0jquazy63WO97rxNef6715+IFlHpTWrzqv8gnXGK0sAAAAp2CwBAACk\nYLMEAACQgs0SAABACjZLAAAAKXJfDbf6ZqeVwc2RijCn8ium726/iiAmVoXmVTmM7PPn7pwdq8Jw\n5lVkbIb7GLPwdj++8yG/GsprL7P5USrkWtnOrVuSwRl+1ejm/f6bNy+XP76tK/ncXXbXP7tjt/T4\nNaLF0WTTh1jVW0ysTcvAj6iGy5tup6VVx2V+45DRw36FWz1l+X0T+13jzdFxbrs7tib3cas/R8fF\n/uM65+HqL1lPvLIEAACQgs0SAABACjZLAAAAKdgsAQAApGCzBAAAkCL/1XBbkhUAS/b6FQBeZZok\nt4Is2mMtVm2WoQqt88ps/dtqUeEWM3KwmIhFq/Lu9sNr6Q2XO4vuTlahDXw92xyx6rSFTjVcTN83\n+tz4p29PxkciVXmdbX51jtf/TpL0o1cqWxyaTqHQ6cYXdV+eiB3dlez1llWs8que8xx9xM+rZXOT\n/d6iz/GMvMcqVlEXiy9auiwRW/8l/3orb218DvLKEgAAQAo2SwAAACnYLAEAAKRgswQAAJCiqgPe\nZvaqpLck/UrSiRDCglosCgci4IEAACAASURBVGhV5ARQjpxAHtSiGu73Qwiv12CeuhjZsTgR65j/\nD+7YjkgVQftNGS7Y3u3HIxVrXrWZDmbrO5elZ5x7vZQ53HjkPq78r+fHrlrJ0vKkqXOiFopzkpWj\nHbP9ip2RMb/6JRZf6MR6ez/jjj288Tk3nkXH7A43fnSX0/9OknRe1decgpoiJ5Ys+agbH9iarJTs\navPn8Kq2JGndM8nfH0OvVry0yfFqtn6JcV6PRj+vYj0XPdtu86tdpcUVz1Er/BkOAAAgRbWbpSDp\nn8xsj5mt8AaY2Qoz221mu4+feKfKywFNj5wAypETaHnV/hnu90IIh8zsg5K2m9kPQwg7Th0QQtgg\naYMkzZpVCFVeD2h25ARQjpxAy6vqlaUQwqHSv0ckfVPSNbVYFNCqyAmgHDmBPJjwK0tmdo6k3wgh\nvFX6+A8lfaVmK6ujzUPXufHCBy9x4we+MpSI9d6brcVI7GD18HDlh91ihof9g7Keri7vIF7KwW+n\n9crQNP9QYLE4tQ++tnJOZDUwmDwQWzgYKW6IiLUZGdyazK3upf7cY11rKp4jdr2YWFsG6SOZ5pnK\nmi0nCl0PuvGxrcnDyP2v+j8rO8f8n9ne8ytWxBA75Nz+Wf9Q9Mi+ZNFP7HdH950r3XjxWOXtWwqz\n/DYohQN+y6B19yfv+6ob/PviHYSX/Me141y/Ndm333TDdVXNn+EukvRNM3tvnm+EEPwyM2BqICeA\ncuQEcmHCm6UQwo/F/2IB/z9yAihHTiAveOsAAACAFGyWAAAAUrBZAgAASFGLdie5sfkMv13H/Axz\nxKrKYrzqtFiVQ6ySzVOLKruYNaumdtUbpLYTa52oXymTlVdBNPJkrO2Bz6tM6r7Wz5+B7X51TuXZ\nhlYR+9k1/6pkG5RlS/22U4Mbd7rxi29Pjh/oe8wdO/CjM/0F3nXIjzvtPQqFTnfkga9d4ca75lVe\nrbpt7xtufGjPKxXPsSoSX35FslWSJI0eTOZhrLWM+itfR63wyhIAAEAKNksAAAAp2CwBAACkYLME\nAACQgs0SAABACqrhTlEsjrjxhb3JWN9X/Kq33nsLbrx/a+X922Ky9ALqUsbqodmRdT/hzeP31sPU\nsW3bC4nY2kj/tlglW6xvVi14c3fL71WFqWPLfS9GbvmrRKT/sP8zccYKf4bB0WQfze4zF7ljY30K\nYxWbR3cl+7qNHn7KHTs6dNSNH062OI3qjvRkW/RBv5Ktqy1Zybbmh1f5k5sfHjqSrARc5xckTgpe\nWQIAAEjBZgkAACAFmyUAAIAUbJYAAABScMB7gjoiB+Bi7U5irUrq2ZYky/U6Z/vjvcPjm29t/FvN\no7n4xRCRA50ZxQ6/erIcEl/3jN/WJHa9RV/xD7PGW1Kg2fUf9uM9Fydj4xv8n+VzLkwe5Jaknv88\nkLxeh//8bLvS/zk8Muyfwu68KXnYfFGkIc8i+QfTa2HoQj+Htu1LPlaFDZ91x/Y+5D9++uPk4ft1\n9/u5Ofrv5rnxWJFWLfDKEgAAQAo2SwAAACnYLAEAAKRgswQAAJDitJslM9toZkfMbP8psQ+Y2XYz\ne6n0r3/aGcghcgIoR04g7yqphtskaZ2kJ06J3SPpOyGEB8zsntLnX6z98ppD+8eS/U7Gntnijo22\nDalBu5OYgYfX12/uwW1O9MN1u16L2KQpnhONVovWKKtu8NudDGyPVKQeTVY3nTTln/+eTWqBnOg5\nFqlwfDkZ6jjX/xnf4bT2kPzKrd6bnV5ZkpTsXpLq3YPJ2NiwX61XC21d/u+xDmcdklRw1jJjrv/4\n9d3t9zDp+tr8RGz5Bj9nV95av6q3mNO+shRC2CHp5+8LXy/p8dLHj0v6ZI3XBTQtcgIoR04g7yZ6\nZumiEMJ7W8nDki6q0XqAVkVOAOXICeRG1Qe8QwhBUojdbmYrzGy3me0+fuKdai8HND1yAihHTqDV\nTXSz9JqZFSSp9O+R2MAQwoYQwoIQwoLp02ZO8HJA0yMngHLkBHJjopulZyXdUvr4Fknfqs1ygJZF\nTgDlyAnkxmmr4czsKUmLJV1gZgcl3SfpAUlPm1mvpJ9IurGei5xsXkVY5d2rTupZ6n9FcU6kOsNx\n9Pk+N56lv5zX602SdMCrepNGR72Ki6ldDURO1JdX+Rbr31bP3nCxPo9T/fnvaZWcGJnh/wwtPJ+s\nxBqf4f9sbvdbnqn9ieTcB7TTH+tUWKfJUvEc60Pqif7uiFTrxX5/DAwn19fzWf8+fnqfP8e4cx9n\nfGyRO7a3t9uN9/U95sZr4bSbpRDCTZGbPlHjtQAtgZwAypETyDvewRsAACAFmyUAAIAUbJYAAABS\nsFkCAABIUUlvuCnPqwibo2w9IWOVNe1OIU6s8iFW5dBxe7JiYHSjX4URmztWrdfRkewRNLLDHQrU\nRKw6LctYr0oua2+4zit7Ilc9VNHa0HwGN/rxZXOTse5r/Z+3xTZ/jvHh5PjOjkjV23Y/HNPTEali\ndhdS+dBOPyXiIuue83Ly98TQNL/6dOeTfgXr2ceHksGX/Mfv22Gxv5A64pUlAACAFGyWAAAAUrBZ\nAgAASMFmCQAAIAUHvCdo0VcibUqODrjh2NvKd3/MicXeUj5yOLv7Y13J4O3+8mIHv6OtHaYlD+4B\nzaKe7U6OPuKPl86r+JpoLr2frrzNSOzgf3GW/zO+9yGnIGCXP/fRXZF+Ii1q0dJliVj/Pr+FVmeb\n36qkqy35eLdf7fxuk1T84VVufGjP07ElVo1XlgAAAFKwWQIAAEjBZgkAACAFmyUAAIAUbJYAAABS\nUA1XAa/lx857/ZP+i+723yK/508jVRhjyQqAkX39/hyRliTeHAX572M//0+zvb99e1vy/mx+9JVM\ncwCeLG1Nss6RpUouNnZkzM9lTG1e+ytJ6tuQ/LmdpfqumcSq0LJU8cUep1h9tXfNWAXr+O/2RWZZ\nfPqFTRCvLAEAAKRgswQAAJCCzRIAAEAKNksAAAApTrtZMrONZnbEzPafEvuymR0ys72l/yK9P4D8\nISeAcuQE8q6SarhNktZJeuJ98YdCCGtrvqIm1HYieTfbbnD6AEkqzvGrH6J93SJ94Dyx/m3DW5OV\nAVn7y3XcvsiNjz7s9ZK7zl/g1LFJUzwn5l9142Qvoe6WX+H/bh8Y+ucGr6QlbNIUyYmFc/1q5cPP\nJ38nNHsPuNHDfiVoR6SnXaxKzjPj7Ufc+MjY/3Dj3ZEKbo9XoS5JIzsqniKz076yFELYIenn9VsC\n0FrICaAcOYG8q+bM0ioz+9fSy6/tNVsR0LrICaAcOYFcmOhm6euSLpc0T1JR0oOxgWa2wsx2m9nu\n4yfemeDlgKZHTgDlyAnkxoQ2SyGE10IIvwoh/FrSY5KuSRm7IYSwIISwYPq0mRNdJ9DUyAmgHDmB\nPJlQuxMzK4QQ3jtt/ClJ+9PGt7qx4eWJ2HjkQPT6jf7bsC/s2uzG+772RiLWNa/bX8e0L7jx0VnJ\ng99tz/tz6Gq/TcvoPv/wuNdiZWex0x1bLI7415wCplpOLLntbxOxwa9ka6XT7Nq6/EOknbOeS8RG\ndiyu61paUV5zYtu2F9x4z7mt91fGjourbzlUT6tu8H+m9Kvydka1ctrNkpk9pZMNVy4ws4OS7pO0\n2MzmSQqSXpV0Rx3XCDQVcgIoR04g7067WQoh3OSEY13sgNwjJ4By5ATyjnfwBgAASMFmCQAAIAWb\nJQAAgBQTqobLq1gLh56bL0nE+iNVb7F6oDny26MUbpufjM3y31K+ECkAGDmRjHXO9ivnsvJarCwq\nfNUdu7k45dugTBnzTySf6QOTsI56Gtzq36M5P0vGpm4d6NQTrfo9r37X9NqSTEYlW6x9S5Y2KLEK\ntyw6rvQrVeuJV5YAAABSsFkCAABIwWYJAAAgBZslAACAFGyWAAAAUlANd4reP9vjxvsfrvyNaLvv\nXOnGCwf8nmyrnx5KxDqO7XTHen3aJGm8K1lRp5/5vd76t/oldbG5h4fHKx57wPxqwqE9T7txtLBd\nM5xg8rlSKx2z/Qqa0YOjFc/R2eY/b0fG/JyIxTsvdCpbjxyqeB3Ip3pWp9Vrbq/KbiLXW/dMMg+7\nbvb7k/ZFKsnfnZ7st1qLyrla4ZUlAACAFGyWAAAAUrBZAgAASMFmCQAAIAWbJQAAgBRUw52iMOZX\n83gVbgMPr882ebtfGSD1Z5unTvq+4lfPLbw9WfXk9YuTpDkL/sqND+358MQXhqZUHPWqZSqvTGsF\nseq5hUuTubyu78l6Lwc1UIveZjHeHLHrNYusVW+x6rl4V9Sk3nv9vm7r7s+0lIbjlSUAAIAUbJYA\nAABSsFkCAABIcdrNkpldambfNbMRM/uBmX2uFP+AmW03s5dK/7bXf7nA5CMngHLkBPKukgPeJyR9\nPoQwZGbnStpjZtsl3SrpOyGEB8zsHkn3SPpi/ZZaO/Ov8ttyjOz7rBv3Wn7E2ppEHR1wwx0dzmG3\nXZH2C5GD1brQPzCXRe9DTgsHyV13rGVKx+3Vr6NF5C4nYgqFTjfe1pX8Xo/s91v6xA5K10KsJUkW\nsfUd+JlffNExnDzM2vnx59yxIzsWT3BVLSd3ORF/bp3Z0HXkTXHOksgtlReIFGbFDuS/knk9lTrt\nK0shhGIIYaj08Vs6eY8ukXS9pMdLwx6X9Ml6LRJoJuQEUI6cQN5lOrNkZpdJ6pL0fUkXhRDee6nj\nsKSLaroyoAWQE0A5cgJ5VPFmycxmSdos6a4Qwpun3hZCCJJC5OtWmNluM9t9/MQ7VS0WaCbkBFCO\nnEBeVbRZMrPpOpkA/SGELaXwa2ZWKN1ekHTE+9oQwoYQwoIQwoLp02bWYs3ApCMngHLkBPKskmo4\nk9QnaTSE8NVTbnpW0i2lj2+R9K3aLw9oPuQEUI6cQN5VUg33UUk3S9pnZntLsS9JekDS02bWK+kn\nkvwSsyY0fo5f9ZZF1nYnPUurrwjqvNKvWBvyqgii7VX63OjIPr/yZ3BjMua1QJGk4Y073Xih8OeJ\nWLE4EllfS8hdTsQs+o/f9G84cVVjFxJRz0q73k/3uvGik24rb/fHrtxRv+qcJtPSObHumeQ3tecy\nv/3VwI+ohnu/VTckK0RXv+T/vhp8yZ+juy1Skd0kTrtZCiH8iySL3PyJ2i4HaH7kBFCOnEDe8Q7e\nAAAAKdgsAQAApGCzBAAAkILNEgAAQIpKquFyx+3HJkknIr3XMujq8ivFaqHvbr9iretr85PBSC+6\nWFXeUKS/3LK5yVj77Njj5McP2BXJka1dDTdl9HT4VZXFJ5Lfa68iRpIGtvtVRbHeW8uviPWOqlzH\n7ORaRg/6vadi6xh5xp97+RXJHCqM+fcRrWvLS0ORWxa7Ub+irjZrab862Qvt6K7h2kyeQcfFlVef\nLrvR+b0kadvf/afIV7wwgRU1Dq8sAQAApGCzBAAAkILNEgAAQAo2SwAAACmm5gHvKyMHvIf9A8od\nty9KxGYM+4f/Omf7c48cjMzdlTwEV1joH3A9cOcWN57Fzof8g6jb5vn3Z47z1vS98u9jrB3LAZ2f\niA3tiSwQTWXnvdvceNcNyYPfA1v9A9RZtXUln19bpq12x24e+rAb71+aXN9CZ15J6ojkfYy3vqOP\nxHLzvExzo3ksvN2Pf9vvGOUWOOzc6j8vRt866sZ7H4q1/Ej+3G6/2i+oqKdYkdEyJ1a4NnkoXZKG\n9jztxhd98JKJLqsheGUJAAAgBZslAACAFGyWAAAAUrBZAgAASMFmCQAAIMWUrIYb3bjTjcdagax+\nOlkptub3IxV17X57iOGtfglFh1MwUDzmv419rIKiS5W/BX3Mwrl+lcMfLXOqoSKtVGLxsWle3K9i\nQnOZ87BX5yKNPVx9a6DOtuqft1mMZax6i7VH8WqQCp+NVCbddSjTNdE8hoeztbDx2p3Mz9j96ugj\n/nNu9LDfkqdZeO1YRp6PlA1G2sVkMfDw+sgt11U9dwyvLAEAAKRgswQAAJCCzRIAAEAKNksAAAAp\nTrtZMrNLzey7ZjZiZj8ws8+V4l82s0Nmtrf0n9+jA8gZcgIoR04g7yqphjsh6fMhhCEzO1fSHjPb\nXrrtoRDC2votrz66uvwShVj/tmU3Jvu36WfVVwPFxKr1eu/1q4eKs5ySuhN+dc6ipf59H+p60F9M\nZJ4sor34WlfuciJm5a2vuPH+3mS/xNgzJVZVlsXoaCzf/KrKwa3JCsyO2f4Ko1VvkfFeb7g1T34g\nsj6/D1YOtUROxKvKkt/r2O+JzX4bTXXfnPy5HesVemCrX2lX7PInP5ChMi+27lqYEVvH1cn4N6pv\nZdpUTrtZCiEUJRVLH79lZqOSmrvjHVBH5ARQjpxA3mU6s2Rml0nqkvT9UmiVmf2rmW00s/Yarw1o\neuQEUI6cQB5VvFkys1mSNku6K4TwpqSvS7pc0jyd/D8K9+84ZrbCzHab2e7jJ96pwZKB5kBOAOXI\nCeRVRZslM5uukwnQH0LYIkkhhNdCCL8KIfxa0mOSrvG+NoSwIYSwIISwYPq0mbVaNzCpyAmgHDmB\nPKukGs4k9UkaDSF89ZT4qSfXPiVpf+2XBzQfcgIoR04g7yqphvuopJsl7TOzvaXYlyTdZGbzJAVJ\nr0q6oy4rrIPBjX584e3Vz913t99jbYZTsCZJBaeSbcusSLlFpO+cZ2iaX+Ez/+ps1W3e/em916/w\nKM7xq4LXRCqqWljuciKrASUrbrqdKjEpXiWXhVuRKmlkhz9+4dLKcyW2Pq/qTZK2TFudiA3tmfK9\nDpsqJwqFTjfece6bbnzo1WQsVskWs7pvcYbR/vMlVmmXRS3myH5Nrwfi4rpdL2ulYi1UUg33L5LM\nucnpsArkHzkBlCMnkHe8gzcAAEAKNksAAAAp2CwBAACkqOSAd+58Oyz2431ZZjnPjc6/6r+48aE9\nftuDA1+7KhEb2eMfiF62wztEF+OvT4rNERu/OBH59l/Grpm7g9yI6Ot7LBmLjO38+HNuvOOY3zph\nzjPJg7X9226KzD7iRs+enTyE3fcVv2XKjBX+Qd7Nj8YObU/5w9xNr1j0nxftX/SP869yYv2jsdYo\nqIfu/5xsUbTzIf9nxIGbGt9Ci1eWAAAAUrBZAgAASMFmCQAAIAWbJQAAgBRslgAAAFJYCKFxFzP7\nmaSflD69QNLrDbv45OA+Npd/H0K4cLIXcSpyIpda6T6SE5OP+9hc3Jxo6Gap7MJmu0MICybl4g3C\nfUQWU+Gx5D4ii6nwWHIfWwN/hgMAAEjBZgkAACDFZG6WNkzitRuF+4gspsJjyX1EFlPhseQ+toBJ\nO7MEAADQCvgzHAAAQIqGb5bM7Doze9HMXjazexp9/Xoxs41mdsTM9p8S+4CZbTezl0r/tk/mGqth\nZpea2XfNbMTMfmBmnyvFc3MfJws50ZrIifohJ1pTnnOioZslMztD0npJ/0FSp6SbzKyzkWuoo02S\nrntf7B5J3wkhzJX0ndLnreqEpM+HEDol/a6klaXvXZ7uY8OREy39fCEn6oCcaOnnS25zotGvLF0j\n6eUQwo9DCL+U9PeSrm/wGuoihLBD0s/fF75e0uOljx+X9MmGLqqGQgjFEMJQ6eO3JI1KukQ5uo+T\nhJxoUeRE3ZATLSrPOdHozdIlkn56yucHS7G8uiiEUCx9fFjSRZO5mFoxs8skdUn6vnJ6HxuInMgB\ncqKmyIkcyFtOcMC7QcLJssOWLz00s1mSNku6K4Tw5qm35eU+ojHy8nwhJ1AreXm+5DEnGr1ZOiTp\n0lM+n12K5dVrZlaQpNK/RyZ5PVUxs+k6mQD9IYQtpXCu7uMkICdaGDlRF+REC8trTjR6s7RL0lwz\nm2NmZ0r6E0nPNngNjfSspFtKH98i6VuTuJaqmJlJ6pM0GkL46ik35eY+ThJyokWRE3VDTrSoPOdE\nw9+U0syWSPrvks6QtDGE8NcNXUCdmNlTkhbrZHfl1yTdJ+l/SXpa0m/qZBftG0MI7z/c1xLM7Pck\nPS9pn6Rfl8Jf0sm/R+fiPk4WcqI1ny/kRP2QE635fMlzTvAO3gAAACk44A0AAJCCzRIAAEAKNksA\nAAAp2CwBAACkYLMEAACQgs0SAABACjZLAAAAKdgsAQAApGCzBAAAkILNEgAAQAo2SwAAACnYLAEA\nAKRgswQAAJCCzRIAAEAKNksAAAApqtosmdl1Zvaimb1sZvfUalFAqyIngHLkBPLAQggT+0KzMyT9\nSNK1kg5K2iXpphDCSOxrpk+fGc466/wJXQ+o1ttvH349hHBhveYnJ7I5+9y33Lj9+jcTsXfePlr1\n9Wae0+7Gz/93r7vx4r+dUfU1mx05AZSL5cS0Kua8RtLLIYQfS5KZ/b2k6yVFk+Css87X78y7rYpL\nAhM3+ML9P6nzJciJDDo//pwbn/H2I4nY0J6nq77e/KtudONLbvtbN75m1XlVX7PZkRNAuVhOVPNn\nuEsk/fSUzw+WYsBURU4A5cgJ5ELdD3ib2Qoz221mu4+feKfelwOaHjkBlCMn0Oyq2SwdknTpKZ/P\nLsXKhBA2hBAWhBAWTJ82s4rLAU2PnADKkRPIhWo2S7skzTWzOWZ2pqQ/kfRsbZYFtCRyAihHTiAX\nJnzAO4RwwsxWSfpHSWdI2hhC+EHNVga0mKmUE95h6WXzPuuObR9e48bb5t5c+QXnZRib1fBKN9zf\n6w/fMtyTiG0euq6WK8qNqZQTyLdqquEUQtgmaVuN1gK0PHICKEdOIA94B28AAIAUbJYAAABSsFkC\nAABIUdWZJQD51n3mL91477yznGifO3ZMRT8+7Mc9owdH3fjI2AE3vvyKJRXPndWyrv5ErDtyH7cN\nr3bj3w6La7kkIJdi7/I/smOxG+/t/Ywb7+t7rOq18MoSAABACjZLAAAAKdgsAQAApGCzBAAAkILN\nEgAAQAqq4QBoS8/lbvzdGYvc+ODWgUSsY3ZHTdd0qoVLu/14ZHyWSrusssz96dv9CsGOZ76QiK19\n8yMTXhPQKpbf8aJ/w65kZWvPskjPoUtn+PHx5M8lKVanmw2vLAEAAKRgswQAAJCCzRIAAEAKNksA\nAAAp2CwBAACkoBoOmGJWr3szEXt3uPKqt8lQz+q2eoo9fgtvWJuILZ/m95Hb/OiHa7om4D3rN/lV\nsH1fu8qN9/7ZHje+eUWyd2PPZePu2OFv+pWt3dd6FW7+HN+e4c9Rix5wMbyyBAAAkILNEgAAQAo2\nSwAAACmqOrNkZq9KekvSrySdCCEsqMWigFZFTgDlyAnkQS0OeP9+COH1GswD5AU5AZQjJ9DSqIYD\nppiO4ZWJWKzaLNbvbfRgsvoFSbHHz3u8l3WtccfuLPgVSMXiyMQXhtzq7f1MxWO33dbjxlc/5Feh\n9d3W78bbu5LjH5zl928b+dF5bnzgp88lYh2j/jraupa48flX3ejGh/Y87cazqPbMUpD0T2a2x8xW\nVL0aoPWRE0A5cgItr9pXln4vhHDIzD4oabuZ/TCEsOPUAaXkWCFJZ57l7yiBHCEngHLkBFpeVa8s\nhRAOlf49Iumbkq5xxmwIISwIISyYPm1mNZcDmh45AZQjJ5AHE94smdk5Znbuex9L+kNJ+2u1MKDV\nkBNAOXICeVHNn+EukvRNM3tvnm+EEP6hJqsCWlNT5cQXzvvfbnxs+A8SsdiB7YVL/bYC4oB3VbzH\ne2FXwR3b+xffc+NrVrXEn6uaKifyZPkdL/o3nPDDw3vfSMTGP+aP3XnvNjfe+5B/IHzZXYcSsUKh\n059cfmHC0Zc+mwwWvuqOHRte7sbnRK44pOsit1RuwpulEMKPJX2k6hUAOUFOAOXICeQF7+ANAACQ\ngs0SAABACjZLAAAAKdgsAQAApKDdCZBTc37r85O9BNTA/J/5rWiklqiGQw14lWWjo8+5Y5fd+KAb\n/6Mrk6133t3X547d0rHaja91qt5isrbjKXzoikSs47ZF7tjRjTvdeE9HspWTJLV1Jat6+/oey7A6\nXlkCAABIxWYJAAAgBZslAACAFGyWAAAAUrBZAgAASEE1HJBTsX5vHbOTVTGTcb1arC82R7Xz1mru\nLHPEFLf6FT7SP1c9N/Jn+M+G3PiWjyXjIzti1W0fruGKKjN+TrI3XGFWrzt2yyz/PvaPrnfjo7OS\nVX+FgtOLLgWvLAEAAKRgswQAAJCCzRIAAEAKNksAAAApOOANSdLy+f/gxgsHkwfs1h2p/C3vUX+9\nvZ9x4ws17sbHhpPtM+p16FvKfsi5FoeiGznvROZeuDTZfsH7vkxkbuSP1zqkWFzsjo02GdlRs+XU\nxbIb5ydixWPD7tiFc/vd+PDeNyKzJw9zL1nyUXfklv/pz8ArSwAAACnYLAEAAKRgswQAAJCCzRIA\nAECK026WzGyjmR0xs/2nxD5gZtvN7KXSv+31XSbQPMgJoBw5gbyrpBpuk6R1kp44JXaPpO+EEB4w\ns3tKn3+x9stDJdZvujwRKxzY5o4d+HqyCkeSRvb78aM6kIh94Ty/Omftmx+JLbFqa3qfc+Or+xbX\n7ZopNqmJcmLwpR433j03+Rb/MW1dhUzX9Krn6tnupBYVYbVqSVKL9cUq3zxe5Zwkret7MtM162yT\nmign0HoKs7oSsVg1XFYdHcmfb4Oj/s/NmNO+shRC2CHp5+8LXy/p8dLHj0v6ZKarAi2MnADKkRPI\nu4meWboohPDe/xodlnRRjdYDtCpyAihHTiA3qj7gHUIIkkLsdjNbYWa7zWz38RPvVHs5oOmRE0A5\ncgKtbqKbpdfMrCBJpX+PxAaGEDaEEBaEEBZMnzZzgpcDmh45AZQjJ5AbE90sPSvpltLHt0j6Vm2W\nA7QscgIoR04gN05bDWdmT0laLOkCMzso6T5JD0h62sx6Jf1E0o31XGSezb8q+dAtm5fsYyNJheeT\nvXMkacbzixKxdU/61Tmrbpjhxrt2+X3E+l9Njp/zW593x2ron/14Bl5lnyTN+Js3q567VpotJ1be\nnuzfJ0l63g97lVsLj037IQAAEV5JREFUI9VwWfqV1ararFl6odWzX55XfTi4dcAdG/vedH78OTc+\nsmPxBFc1cc2WE2g9Aw+vT8R6Ola6Y0dm+JVsw3rEjY+OJn+OxfrL9en/cuOn3SyFEG6K3PSJ030t\nkEfkBFCOnEDe8Q7eAAAAKdgsAQAApGCzBAAAkILNEgAAQIpKesMhg0Kh042v/+Ibbrz4xPs7BEij\nW/2T/p3X+pVsGt+ZCK26wa/kWfeMX2l09vEhf+7pyUq7o12r/bFDi92w95gsKnzVHdv2vF+h8OC5\nV7nxzo8nKwcnoxpoMnlVJJK0rCv5vYvJ0qusVrJUz9Wiv1xWtZi7nhV1S/b6fWlH6nZFoH6673Qq\n38b8Ku2xe/3nfvGc/W589f3JiuG+r/m/k2N4ZQkAACAFmyUAAIAUbJYAAABSsFkCAABIwQHvCiy/\n48VErKej2x1bfOJDbnzgb/zDZN3XHkjEZhzzD5aueyZywDuD+MFvf3zPZckDdv1P+geH18//Azfu\nHZTtOubP0dP3pL+QiELBaw0ztY64ugcjpWi7E1THa1WiyGFwd6wm50A90Ax6ez/jxoecn1d9fY+5\nY1evm+PPre9FrtqVvN6epyNjfbyyBAAAkILNEgAAQAo2SwAAACnYLAEAAKRgswQAAJBiSlbDLZ//\nD2786H6/wq17V7LaaN3fbXPHrrrBr1grdPhxv/1I9VVvWXlVb5LUfnWyikCv+pU/B15+0I2ve/Mj\nyeCR89yxa3qf89cxvMaNFzqS1YRrfnijO3bwhfvdeKvbvML/fvTeXHm7k5hYNVes+qsVZW2l0uh1\nxCrnum5Y60+UsaIUqIfYz/LO2cnqckla8//+l0Rs/abL3bHFY8NuvO+vf9eNL1nyUSf6ijs2hleW\nAAAAUrBZAgAASMFmCQAAIAWbJQAAgBSn3SyZ2UYzO2Jm+0+JfdnMDpnZ3tJ/S+q7TKB5kBNAOXIC\neVdJNdwmSeskPfG++EMhhEg5RjZer5g56nHHDm705+g4t92NL1q6LBFb94xf9Rbrm3Z01xZnrFMl\nlsKvemt+R3clqw56LvPHDo/5PwvX/5b/uHo2P/lLN77qhmTVW8zqP/4rfx0PVzzF6WxSnXMii96b\nezON96quYlVvg1sHJrSmPPOq02KVbDFepV3WOZrse7NJTZQTqI/5V/mVxoWhZI/O2O/qzof837/L\n5iXnKB6bX/niFKt6q43TvrIUQtgh6ed1WwHQYsgJoBw5gbyr5szSKjP719LLr/7LOpLMbIWZ7Taz\n3cdPvFPF5YCmR04A5cgJ5MJEN0tfl3S5pHmSipL8dyOUFELYEEJYEEJYMH3azAleDmh65ARQjpxA\nbkxosxRCeC2E8KsQwq8lPSbpmtouC2gt5ARQjpxAnkyo3YmZFUII751y/JSk/Wnj39N+zhtuq5FF\nz7+ZHHu1f2i189MVLzMqdpA7xm35EeEdiJbih6LrtY6sB8pjc2ebxz+EPTJW+eHsrLz1rfp/6na5\nqInmRBaxw5UxsTYZWcbGvnfLr6i+sKkW7USyHoqu1zqibWFqMEfse7NwqX9Qdl2TtDtpRE6gPryi\nK0la9Lz/4mD7Q8mCrJ33+i3Blt11KHLVxZUsTVJ8fWPTvuDGdz77qYrnjjntZsnMntLJe3GBmR2U\ndJ+kxWY2T1KQ9KqkO6peCdAiyAmgHDmBvDvtZimEcJMT7qvDWoCWQE4A5cgJ5B3v4A0AAJCCzRIA\nAEAKNksAAAApJlQNN1HHjl+sncU/T8RH5z6SiHWM7nTnGB+e4caXzfXfFr3/VX+8p7NtTsVj4/zq\nnCxVO7GqmHczrKL35kVVzyFJy69IPia1qB6qZ4Xc6u9WXgXWSpbc9rf+DcMrM83jPb+yVM5NdW6r\nkoxz1KKKD6hWrML28Ea/3di2rnE3Pufe5M/ztW9+ZOILq7FicaTqOXhlCQAAIAWbJQAAgBRslgAA\nAFKwWQIAAEjBZgkAACBFQ6vhYkZ2LE7GMs7x7R/5ca+HTKx/THHXgBvvWepUybX7fZniKq8gG9m3\n2o0PD/uVCN13JquhBgb9vjyjo5E+U3P7/WvaBxKx4hnnu2NjCh+6Ihm89KOZ5ljW/3TkljOToR2L\nI2MHM12z1cWqFhdm6GOWpUK0FlWSWcWumaXaLDY2y/3Jet+phkMzKAx91o333uv/jOj7iv/7Y21o\njsq37oV+z8rNj75S9dy8sgQAAJCCzRIAAEAKNksAAAAp2CwBAACkYLMEAACQoqHVcMd/+W7FPVoK\nhU4/7lVWpdi27YVErFj8cOSan3LjOzdkumTVCh9K9sqTJJkf7vtaMtY1z6/WWzjXn2PwJb8XUMeC\nZFXEkisrr6Y66XsVjyzM6nLjS4696cZH9yWrM9pOrHXHDr5wf8XraEZ9f/27bnytXwDSsupVKZa1\ncq4WVXIxsf6PaB6x30FLlviVvF6V9eZH/d81k2HLfS8mg+3+z/2d9/rV1N8Oi2u4oto7+nxf5JbF\nVc/NK0sAAAAp2CwBAACkYLMEAACQ4rSbJTO71My+a2YjZvYDM/tcKf4BM9tuZi+V/m2v/3KByUdO\nAOXICeRdJQe8T0j6fAhhyMzOlbTHzLZLulXSd0IID5jZPZLukfTFWi0sdhC80gPiaWIH96LjMxwq\n75qXrRVIs4i2O9n9RiJ2YHe2ub3HZHhvct7Y2JPjr6r4ekN7Hqt47ARNSk6svv8X/g3PZ5tnbNg5\nFB85cBw7al3P1ibe3FkPfddijmZRi7Y1DTApOVFP7XP9QptCl1+wMravnqup3PyrbnTjR5/620Ss\n/eoZ7ti1bzZH+5KsOq/0D6xLh6qe+7SvLIUQiiGEodLHb+lkk7NLJF0v6fHSsMclfbLq1QAtgJwA\nypETyLtMZ5bM7DJJXZK+L+miEMJ7/4t6WNJFNV0Z0ALICaAcOYE8qnizZGazJG2WdFcIoexNb0II\nQVKIfN0KM9ttZruPn3inqsUCzYScAMqRE8irijZLZjZdJxOgP4SwpRR+zcwKpdsLko54XxtC2BBC\nWBBCWDB92sxarBmYdOQEUI6cQJ5VUg1nkvokjYYQvnrKTc9KuqX08S2SvlX75QHNh5wAypETyLtK\nquE+KulmSfvMbG8p9iVJD0h62sx6Jf1Ekn8EvwllrajLMn5oT9bV5F+Wx6RFHr9JyYkDd25x4103\nLHLjI2MH3HiWqrDi7f41Fw6vrHgOr/ouq6ytSupZreddM1ZNePb4Tjf+rhOrxeMk+dW+tagiPo3c\n/Z4Y2bE4Eo99xXl1Wkk2K99KVr1J0vBY8nn77gy/LZb0Sg1X1DhHH4nlffXfm9NulkII/6JoVzJ9\nouoVAC2GnADKkRPIO97BGwAAIAWbJQAAgBRslgAAAFKwWQIAAEhRSTUcgCYWq6LqbJtT9dxrVvlV\nJFvuW52IDXzdr6yJVazFKsgGtw5UPEdMlvHR3mtLY5VCSbGqt6GFw268Y9ivYKwFr4fgylvrdrkp\np++333Tj/a8m+6wt3+A/D1feWn212Zre59z4DPW58e4rk3m17K6699FsqParu/wbflT9480rSwAA\nACnYLAEAAKRgswQAAJCCzRIAAEAKDngDLS52UFp1bPmx7C8/7EQP+YOPROJDlV+vcMb5lQ/OKjL3\nur4nq5+73z8g399b/dQxflucj9TvgjXWfuG4lt/xYiI+OposZFg4t9+dY3jvG268+H/2+/EM7WBG\nD/tthHpvXpuI9a3wD1sXCvMyrcN7PNp3rXHHjh9b78Z7+pqjHUur4pUlAACAFGyWAAAAUrBZAgAA\nSMFmCQAAIAWbJQAAgBRUwwEtov+cm9x45A3+M7cIaWZZqpVagdfSJdpeJWNVY9sNR5NBvyirKZ07\n63x1L1ySiPdcvC05uN2JSTq80a+Si9SNqve+2C2O9uTaJKnv7mSJY3vXuDt29Z1+bhbGLvGvuevy\nRGjnyw+6Qwfm/rU/h56OxPNjZEbsib646rl5ZQkAACAFmyUAAIAUbJYAAABSsFkCAABIcdrNkpld\nambfNbMRM/uBmX2uFP+ymR0ys72l//xTb/j/2rufEKvKOIzjz4MaEhYZhYpZSrhIWqiYhBYM0sLc\nWC4iF2EU1MIowRYigrRombWpyHDKwIrAoYRWZkG2kVQkrcE/ZX+MSYsSlf6avxb3LGa855zmOvee\nM+f1+9nMvb97x/O+d84jP53znheJIRPASGQCqRvNariLktZHxEHb10k6YHt39tqLEdG+IQ6Qtloy\nUbQi7IVDd+bWlxSsospbJfd2f9FmZX2jGRo6tGxje+2Pw/nvLVrV+Ofe/D3ANh3ru8JRjUnXMvH9\nyb+09tGvc17pYD/CgvN2xox5ufXpp5a21c5OfDb3vcvmTM6tr3ijPUO/7c1fnZW/f580efqc3PrQ\nve0bKX4895Xc9x48kP6qt6K9/xZG/n550kdjPub/NksRMSRpKHt83vagpIL1jUD6yAQwEplA6jq6\nZsn2bLVu67IvKz1l+wvb/bandnlswLhHJoCRyARSNOpmyfYUSTslrYuIc5JelXS7pPlq/Ysi9w5Z\ntp+wvd/2/n8u/t6FIQPjA5kARiITSNWomiXbk9QKwI6IGJCkiDgdEf9GxCVJr0tanPe9EbE1IhZF\nxKJJE6/t1riBWpEJYCQygZSNZjWc1bpZ/mBEbBlWH35/+AclHen+8IDxh0wAI5EJpG40q+GWSnpE\n0mHbh7LaRkmrbc+XFJK+lfRkT0YIjD/jKhNFq18eX3hfbv2GBe37YK26sDD3vR8eu/JxodiOt9pX\nSa1akL83XN7PS5IGJg7l/+HHrr/icY3BuMpEkaIVpdu25dXzVt9JO1/LW6lXpK+D90o6V1DPXeGY\n/qq3IkV/523afLTgG/J/lp0YzWq4zyQ556X8HQyBxJEJYCQygdRxB28AAIASNEsAAAAlaJYAAABK\njOYCbwBN9PTJ3PLghfatFgaOt2+nIKngwlKM1R2PLWmrne0vuGC7yF1dGgyQiFXPjf1C7iL8zxIA\nAEAJmiUAAIASNEsAAAAlaJYAAABK0CwBAACUcERUdzD7Z0nfZU9vkvRLZQevB3McX26LiJvrHsRw\nZCJJTZojmagfcxxfcjNRabM04sD2/ohYVMvBK8Ic0Ymr4bNkjujE1fBZMsdm4NdwAAAAJWiWAAAA\nStTZLG2t8dhVYY7oxNXwWTJHdOJq+CyZYwPUds0SAABAE/BrOAAAgBKVN0u2l9s+avuE7Q1VH79X\nbPfbPmP7yLDajbZ32z6efZ1a5xjHwvYs25/Y/sr2l7afyerJzLEuZKKZyETvkIlmSjkTlTZLtidI\nelnS/ZLmSVpte16VY+ihNyUtv6y2QdKeiJgraU/2vKkuSlofEfMk3S1pbfazS2mOlSMTjT5fyEQP\nkIlGny/JZqLq/1laLOlERHwTEX9LelfSyorH0BMR8amkXy8rr5S0PXu8XdIDlQ6qiyJiKCIOZo/P\nSxqUNFMJzbEmZKKhyETPkImGSjkTVTdLMyX9MOz5qayWqmkRMZQ9/knStDoH0y22Z0taIGmfEp1j\nhchEAshEV5GJBKSWCS7wrki0lh02fumh7SmSdkpaFxHnhr+WyhxRjVTOFzKBbknlfEkxE1U3Sz9K\nmjXs+S1ZLVWnbc+QpOzrmZrHMya2J6kVgB0RMZCVk5pjDchEg5GJniATDZZqJqpulj6XNNf2HNvX\nSHpY0q6Kx1ClXZLWZI/XSPqgxrGMiW1L2iZpMCK2DHspmTnWhEw0FJnoGTLRUClnovKbUtpeIekl\nSRMk9UfE85UOoEdsvyOpT63dlU9L2izpfUnvSbpVrV20H4qIyy/uawTb90jaK+mwpEtZeaNav49O\nYo51IRPNPF/IRO+QiWaeLylngjt4AwAAlOACbwAAgBI0SwAAACVolgAAAErQLAEAAJSgWQIAAChB\nswQAAFCCZgkAAKAEzRIAAECJ/wDEPVu+gsikCAAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 720x720 with 9 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rVvd442yFQDd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# preprocess data for model input\n",
        "X_train = X_train.reshape((-1, 28, 28, 1))\n",
        "X_test = X_test.reshape((-1, 28, 28, 1))\n",
        "\n",
        "y_train = to_categorical(y_train)\n",
        "y_test = to_categorical(y_test)\n",
        "\n",
        "# normalize data so pixel values in range [0,1]\n",
        "X_train = X_train / 255\n",
        "X_test = X_test / 255"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ghECI9mVKXG0",
        "colab_type": "text"
      },
      "source": [
        "# Convolutional Neural Network Set-Up"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "80pT6OstKh8m",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras.layers import Conv2D, MaxPool2D\n",
        "from keras.layers import Flatten, Dense, BatchNormalization\n",
        "from keras.models import Sequential\n",
        "\n",
        "def create_cnn1():\n",
        "  model = Sequential()\n",
        "\n",
        "  model.add(Conv2D(32, (3, 3), activation='relu', padding=\"same\", input_shape=(28,28, 1)))\n",
        "  model.add(BatchNormalization(axis=-1))\n",
        "\n",
        "  model.add(Conv2D(32, (3, 3), activation='relu', padding=\"same\"))\n",
        "  model.add(BatchNormalization(axis=-1))\n",
        "  model.add(MaxPool2D(pool_size=(2, 2)))\n",
        "\n",
        "  model.add(Conv2D(64, (3, 3), activation='relu', padding=\"same\"))\n",
        "  model.add(BatchNormalization(axis=-1))\n",
        "\n",
        "  model.add(Conv2D(64, (3, 3), activation='relu', padding=\"same\"))\n",
        "  model.add(BatchNormalization(axis=-1))\n",
        "  model.add(MaxPool2D(pool_size=(2, 2)))\n",
        "\n",
        "  model.add(Flatten())\n",
        "  model.add(Dense(128, activation='relu'))\n",
        "  model.add(Dense(10, activation='softmax'))\n",
        "\n",
        "  return model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ic6zgMo63r9E",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 595
        },
        "outputId": "64a70e5e-6ee5-46fa-fa22-28a9eef5fbe4"
      },
      "source": [
        "# initialize and create a model summary\n",
        "cnn1 = create_cnn1()\n",
        "cnn1.summary()"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_3\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_9 (Conv2D)            (None, 28, 28, 32)        320       \n",
            "_________________________________________________________________\n",
            "batch_normalization_9 (Batch (None, 28, 28, 32)        128       \n",
            "_________________________________________________________________\n",
            "conv2d_10 (Conv2D)           (None, 28, 28, 32)        9248      \n",
            "_________________________________________________________________\n",
            "batch_normalization_10 (Batc (None, 28, 28, 32)        128       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_5 (MaxPooling2 (None, 14, 14, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_11 (Conv2D)           (None, 14, 14, 64)        18496     \n",
            "_________________________________________________________________\n",
            "batch_normalization_11 (Batc (None, 14, 14, 64)        256       \n",
            "_________________________________________________________________\n",
            "conv2d_12 (Conv2D)           (None, 14, 14, 64)        36928     \n",
            "_________________________________________________________________\n",
            "batch_normalization_12 (Batc (None, 14, 14, 64)        256       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_6 (MaxPooling2 (None, 7, 7, 64)          0         \n",
            "_________________________________________________________________\n",
            "flatten_3 (Flatten)          (None, 3136)              0         \n",
            "_________________________________________________________________\n",
            "dense_5 (Dense)              (None, 128)               401536    \n",
            "_________________________________________________________________\n",
            "dense_6 (Dense)              (None, 10)                1290      \n",
            "=================================================================\n",
            "Total params: 468,586\n",
            "Trainable params: 468,202\n",
            "Non-trainable params: 384\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lbtxkCP64Isc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras import optimizers\n",
        "\n",
        "# compile the model using adam's and categorical crossentropy\n",
        "# use accuracy as the metric to measure model performanc\n",
        "\n",
        "adam = optimizers.adam(lr=0.001)\n",
        "\n",
        "cnn1.compile(loss='categorical_crossentropy',optimizer=adam, metrics=['accuracy'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xSoX7BEB6Ac6",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "outputId": "a52de81d-6100-4022-ae8c-7a1cf78857b4"
      },
      "source": [
        "from keras.callbacks import EarlyStopping\n",
        "\n",
        "callback = [EarlyStopping(monitor='val_loss', min_delta=0, \n",
        "                         patience=5, verbose=0, mode='auto', \n",
        "                         baseline=None, restore_best_weights=False)]\n",
        "\n",
        "#define the number of epochs and batch size\n",
        "epochs = 20\n",
        "batch = 64\n",
        "\n",
        "#fit the model\n",
        "model = cnn1.fit(X_train, y_train, epochs=epochs, batch_size=batch, callbacks=callback, verbose=1, validation_data=(X_test, y_test))\n"
      ],
      "execution_count": 65,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 60000 samples, validate on 10000 samples\n",
            "Epoch 1/20\n",
            "60000/60000 [==============================] - 9s 153us/step - loss: 0.0222 - acc: 0.9920 - val_loss: 0.4263 - val_acc: 0.9225\n",
            "Epoch 2/20\n",
            "60000/60000 [==============================] - 9s 153us/step - loss: 0.0181 - acc: 0.9938 - val_loss: 0.4536 - val_acc: 0.9249\n",
            "Epoch 3/20\n",
            "60000/60000 [==============================] - 9s 155us/step - loss: 0.0198 - acc: 0.9932 - val_loss: 0.4586 - val_acc: 0.9227\n",
            "Epoch 4/20\n",
            "60000/60000 [==============================] - 9s 154us/step - loss: 0.0207 - acc: 0.9930 - val_loss: 0.4781 - val_acc: 0.9256\n",
            "Epoch 5/20\n",
            "60000/60000 [==============================] - 9s 153us/step - loss: 0.0147 - acc: 0.9947 - val_loss: 0.4466 - val_acc: 0.9257\n",
            "Epoch 6/20\n",
            "60000/60000 [==============================] - 9s 153us/step - loss: 0.0188 - acc: 0.9942 - val_loss: 0.4881 - val_acc: 0.9238\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uIHsw8iRKIsi",
        "colab_type": "text"
      },
      "source": [
        "# Model Evaluation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NJ0gYKzhGKEP",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "2d6721f7-2d70-48f7-84f2-594c02aacc40"
      },
      "source": [
        "evaluate = cnn1.evaluate(X_test, y_test, verbose=1)\n",
        "\n",
        "print(\"Model Validation Accuracy: {} %\".format((round(evaluate[1], 4) * 100)))"
      ],
      "execution_count": 66,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "10000/10000 [==============================] - 1s 77us/step\n",
            "Model Validation Accuracy: 92.38 %\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OQj8aapFJe8W",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 404
        },
        "outputId": "b9f242ec-1342-4b50-a58c-dcb0726caba4"
      },
      "source": [
        "#graph model accuracy\n",
        "plt.figure(figsize=(10,6))\n",
        "plt.plot(model.history['acc'])\n",
        "plt.plot(model.history['val_acc'])\n",
        "plt.title('Model Accuracy')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Train', 'Test'], loc='upper left')\n",
        "plt.show()"
      ],
      "execution_count": 71,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmcAAAGDCAYAAABuj7cYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAF8v8D/v6SfCC7aqQAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OtyksO6kTlCg",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 404
        },
        "outputId": "2f8aaa4e-d200-4257-c0b8-e2454ee9633f"
      },
      "source": [
        "# graph model loss\n",
        "plt.figure(figsize=(10,6))\n",
        "plt.plot(model.history['loss'])\n",
        "plt.plot(model.history['val_loss'])\n",
        "plt.title('Model Loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Train', 'Test'], loc='upper left')\n",
        "plt.show()"
      ],
      "execution_count": 70,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmEAAAGDCAYAAABjkcdfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}