mnist-fashion-items/mnist-fashion-item-classification-with-a-fully-connected-neural-network.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "MNIST Fashion Item Classification with a Fully-Connected Neural Network",
      "provenance": [],
      "authorship_tag": "ABX9TyOuU3tHbqQupgCs2PTnk/gz"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aq7MqgGKBfOq"
      },
      "source": [
        "# MNIST Fashion Item Classification with a Fully-Connected Neural Network\n",
        "*By Carlos Santiago Bañón*\n",
        "\n",
        "**Year:** 2020\n",
        "\n",
        "**Technologies:** Python, NumPy, Matplotlib, TensforFlow, Keras\n",
        "\n",
        "**Discipline(s):** Computer Vision, Deep Learning\n",
        "\n",
        "**Keywords:** `classification`, `computer-vision`, `deep-learning`, `fully-connected`, `mnist`, `multi-class-classification`, `neural-network`"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HSmvrCh8CqZq"
      },
      "source": [
        "This notebook presents an MNIST fashion classifier built with a fully-connnected neural network in TensorFlow and Keras."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Uzlnsvf1DGpk"
      },
      "source": [
        "## 1. Import Statements\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l8DhknOWDIc4"
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "%tensorflow_version 2.x\n",
        "import tensorflow as tf"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Yuvx7u4qDQ0p"
      },
      "source": [
        "## 2. Data Preprocessing\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qI70NJ8SN4An"
      },
      "source": [
        "The first step is to preprocess our data. Here, we load the MNIST fashion dataset from the Keras datasets library, split it into training and test sets, and normalize the matrices."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IsaUq1mDDVLT"
      },
      "source": [
        "# Load the MNIST dataset.\n",
        "fashion_mnist = tf.keras.datasets.fashion_mnist\n",
        "train_data, test_data = fashion_mnist.load_data()"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nWuEpda6EBMN"
      },
      "source": [
        "# Divide the data into features and labels.\n",
        "train_images, y_train = train_data\n",
        "test_images, y_test = test_data"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wrAabZnyDinP"
      },
      "source": [
        "# Normalize the images.\n",
        "X_train = train_images.astype('float32') / 255\n",
        "X_test = test_images.astype('float32') / 255"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FVEOuuLPOJVH"
      },
      "source": [
        "Further, the following are the shapes of each matrix, as well as a visualization of a random MNIST fashion item."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9PpPi0qpNI_L",
        "outputId": "b773137c-4ac2-4e48-e4e5-c472653359d6",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "# Show the shapes of the data.\n",
        "print(\"Training Images:\", X_train.shape)\n",
        "print(\"Testing Images:\", X_test.shape)\n",
        "print(\"Training Labels:\", y_train.shape)\n",
        "print(\"Test Labels:\", y_test.shape)"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training Images: (60000, 28, 28)\n",
            "Testing Images: (10000, 28, 28)\n",
            "Training Labels: (60000,)\n",
            "Test Labels: (10000,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mSKjBaeNNc1m",
        "outputId": "d3da9311-64e2-43ea-f8d0-f9cb63e5af9a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        }
      },
      "source": [
        "# Show a sample MNIST fashion item.\n",
        "plt.imshow(train_images[10])\n",
        "plt.show()"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD4CAYAAAAq5pAIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAUD0lEQVR4nO3da4yc1XkH8P+zc/VefFnfWINtjOO0OJAYugGi0JYKlRIaCdJUNEhNqYRqFEEEVT4UUanwpS2qmpB8qCI5BcVpE6JUhEIbN8JxUCmidTHINcZOuRgb2/V6F1/2vjszO08/7DhaYN//Wea+nP9PWu3s+8x558y7++w7M897zjF3h4h89HW0ugMi0hxKdpFIKNlFIqFkF4mEkl0kEulmPljWcp5HVzMfsikslaLxqUtyNL68c4LGR87xY5YZGKfxxaq0ij/v7MppGp8aTT7u2f/7aB6zKYyj4NM2X6ymZDezmwF8C0AKwN+7+yPs/nl04Vq7sZaHbEuppcto/PBfbKHx37vqFRr/6ZPX0fglf/UijS9W737xMzS+8Q/fpPHDP08+7hse/mges72+JzFW9ct4M0sB+DsAnwOwFcAdZra12v2JSGPV8p79GgBvuvsRdy8A+CGAW+vTLRGpt1qS/WIAx+f8fKKy7T3MbLuZ7TOzfUXw91gi0jgN/zTe3Xe4e7+792fAP6gSkcapJdlPAlg/5+dLKttEpA3VkuwvAdhiZpvMLAvgSwCeqU+3RKTerJZRb2Z2C4BvYrb09ri7/yW7/1Lr9cVaenvrB9sSY3+6LbncAQB5K9L4f41spvF71vycxv97alNi7GdnLqdtX357A42XRzM0nl5eoPGvfPL5xNiyFL++YEtugMb3jH6CxjdkzyTGdp/lhaPhr6yh8fKBX9B4q+z1PRjxs/Wvs7v7LgC7atmHiDSHLpcViYSSXSQSSnaRSCjZRSKhZBeJhJJdJBI11dk/rHaus49/8VoaX3PfkcTY0fO9vG33GI13GP8d9OZ4Pfrqpe8kxtZlztG2L4x8nMZ3vXYFjX/+igM0vjKTPG78rYlVtO3hMxfR+K/0DtL42yPJv5f1Pedp24HxpTSeu+kojbcKq7PrzC4SCSW7SCSU7CKRULKLRELJLhIJJbtIJJo6lXQ7O3kjL3+dPvGBGbd+KZvjQ1inSnyYaD7N2795npeopmaSf42hsl62Y4bGr9nyNo2fLfDpngemkktYofLW1WuO0/jQVDeNp8hzP3i6j7Zd1c2nmp7+3U/TeO4nL9F4K+jMLhIJJbtIJJTsIpFQsotEQskuEgklu0gklOwikVCdvaLrIl5XnSDL/4YWupkq8cOcSfFad1eWT9c8VkzuwJkJXgfPpUs0HqrTF8v8fNHXNZIY683zobuhOvrpiR4aL/u8Iz0BAKmOctVtAWDg1/nvdNNPaLgldGYXiYSSXSQSSnaRSCjZRSKhZBeJhJJdJBJKdpFIxFNn70jRcGj88jsj+cTYBIkBQGdgvHtILsVr4fkU2X8n33c+sO/xUpbGl4DX4dOknp1PTdO2GeO18M7APABnpwNPnpgJ1NlTm/n04O2opmQ3s6MARgHMACi5e389OiUi9VePM/tvufu7ddiPiDSQ3rOLRKLWZHcAz5rZy2a2fb47mNl2M9tnZvuK4O/RRKRxan0Zf727nzSzNQB2m9kv3P35uXdw9x0AdgCza73V+HgiUqWazuzufrLyfRDAUwCuqUenRKT+qk52M+sys54LtwHcBOBgvTomIvVVy8v4tQCeMrML+/mBu/+0Lr1qgI4r+dLEqQ5eZ0/nk2u6xRE+oP3cMB9Tng2MKd+8bJjGp2aS56XvzvDPSULj1dOBeeVD7SdInZ5eH7CAfZecn6vYmPTRSX5tRMjlawdonP81tUbVye7uRwB8qo59EZEGUulNJBJKdpFIKNlFIqFkF4mEkl0kEtEMcZ28hE9LPFXgZSBnUybz0ZDoOM7LPEOBaY3Pjy+hcSOPv6xzkrYtBKa5ninzJxdqz6bJPpfjz2smME31ZIEvhT1yOvl33tHJy52d3bxkefR8L433refl2NLxEzTeCDqzi0RCyS4SCSW7SCSU7CKRULKLRELJLhIJJbtIJKKps0+s5k916PQyGu9cOpUYu3/bHtr2m//6eRovD/B6s69NfmwAyJKpqsemeL23UOTHxQNzC5Vn+PmiYMlTeOcyvNY9HejbyBC/duKmq5KnVyiV+dTi/37kYzSe6ebXL4xtW0fjedXZRaRRlOwikVCyi0RCyS4SCSW7SCSU7CKRULKLRCKaOvvkaj4uO9dVoPG//uRTibFP5wZp23/a9ms0PvCfvCa7ZiufSnpoJLneXAiMCe8IjKUvFnk9OpPltfJ0Knn/PTk+ZvzSZWdpfO/JpTQ+NJV8XB7Z+M+0bW+WTwb94uAm/tif4qm1/l9ouCF0ZheJhJJdJBJKdpFIKNlFIqFkF4mEkl0kEkp2kUiYhwYs19FS6/Vr7camPd6HkdrKl3QeezR5zHj3V/n/zNfvXk3j1sfHq/cExk6PjCWPh89k+JLLIaE6PJuzHgBKpeRj09PJ6+yXr+TLIhfKvJY9+vvJy0UffnAjbZvv43X2jX90hMbLExM03ih7fQ9G/Oy8v5Xgmd3MHjezQTM7OGdbr5ntNrM3Kt9X1LPDIlJ/C3kZ/10AN79v2wMA9rj7FgB7Kj+LSBsLJru7Pw/g/dct3gpgZ+X2TgC31blfIlJn1V4bv9bdT1VuDwBYm3RHM9sOYDsA5NFZ5cOJSK1q/jTeZz/hS/yUz913uHu/u/dnwCc/FJHGqTbZT5tZHwBUvvNhXyLSctUm+zMA7qzcvhPA0/Xpjog0SvA9u5k9AeAGAKvM7ASAhwA8AuBHZnYXgGMAbm9kJ5th5tDrNL7kd0jbwL6XH1pD45dde5zGDw700TgrdYcuowjVyTs6+A46jMdT2eQ6/fAony9/ajlffz3bwY986VRynX7LV3kNP4RffdCegsnu7nckhNrz6hgRmZculxWJhJJdJBJKdpFIKNlFIqFkF4lENFNJh2pMluJTJoPEfZoP1Vz1ygiND/5BD427B/pOhqGGhriWSvx5l8uh2hwPp0nfQs/rzFQXjV+/+i0aHwIv3TGWri01vMSn2G4FndlFIqFkF4mEkl0kEkp2kUgo2UUioWQXiYSSXSQS8dTZA2M9g3XRmeqnZE4N82mJQ0LLJudyydNch+roKbKkMhAeIhsa4lomtfRcPrnfAHBugg+BHSuFZj6qfiCqh37fTZyCvV50ZheJhJJdJBJKdpFIKNlFIqFkF4mEkl0kEkp2kUjEU2evkaWTx0Z7sUDbeo6Pq56e4fXgcpH/T053JrefDNTo81leTy7O8PahOnupnNz37jyfB2CywI/bs+/8Ko2vwyEapyxwHvTalsJuBZ3ZRSKhZBeJhJJdJBJKdpFIKNlFIqFkF4mEkl0kEqqzN8HEpctpfLrI55VP56qfg7y7k9eyC6Xa/gTYeHUAyKaT+z5d5I9dy1h5AEh9fHNibOZ1Pue8dfB9+yJcszl4Zjezx81s0MwOztn2sJmdNLP9la9bGttNEanVQl7GfxfAzfNsf9Tdt1W+dtW3WyJSb8Fkd/fnAZxtQl9EpIFq+YDuXjM7UHmZvyLpTma23cz2mdm+Ivj7RxFpnGqT/dsANgPYBuAUgK8n3dHdd7h7v7v3ZxCaIFBEGqWqZHf30+4+4+5lAN8BcE19uyUi9VZVsptZ35wfvwDgYNJ9RaQ9BIusZvYEgBsArDKzEwAeAnCDmW0D4ACOAri7gX1sDzUUVgc+ww9zOlDrzgbGnKfIGuhTgTHhXXk+Fj80pnyGjFcH+Jj1kck8bcvWdg/tGwAKFy9LjKVep02BFB/HjzZcfz0kmOzufsc8mx9rQF9EpIF0uaxIJJTsIpFQsotEQskuEgklu0gkNMR1gYJL+BLFTVP8DiX+P7drCS8x5TPJZaBQ6Y0NQQWAQmDJ51DpjenK8bLf6CS/4jKf5Us+n7k8ubS35jnaFCgvviWZQ3RmF4mEkl0kEkp2kUgo2UUioWQXiYSSXSQSSnaRSKjOfkFHYEhjObnObpksbbpmFZ8qemKat/fAlMk8ynVnahviWprh54sUmQ56KtC2o4PXukNTUY9sSR4iu4a2rO26inalM7tIJJTsIpFQsotEQskuEgklu0gklOwikVCyi0RCdfaKWpboTa3qpW2HzvXQ+EW9vA5/bnwJja/uGk+MDRb5Y7NpqBcineLt2bLLmUBbd17rzqZ5vHvTMI1T5LoKAIAFrm7w9hsPrzO7SCSU7CKRULKLRELJLhIJJbtIJJTsIpFQsotEQnX2C6z6/3uFj/XReE/XJI2HKrKh+dG7MsnzyofGwneTtgDQmeXLKo8HxuKXyeMvy/H59IdKXTQemtO+QMa7W47PSe/T/LhYYElnb8MlnYN/4Wa23syeM7NDZvaamd1X2d5rZrvN7I3K9xWN766IVGshp7MSgK+5+1YA1wG4x8y2AngAwB533wJgT+VnEWlTwWR391Pu/krl9iiAwwAuBnArgJ2Vu+0EcFujOikitftQ79nN7FIAVwHYC2Ctu5+qhAYArE1osx3AdgDIo7PafopIjRb8qZSZdQN4EsD97v6ekRvu7kj4nMndd7h7v7v3Z8A/FBGRxllQsptZBrOJ/n13/3Fl82kz66vE+wAMNqaLIlIPwZfxZmYAHgNw2N2/MSf0DIA7ATxS+f50Q3q4CJz5BC9Pre3h/wdPDi+j8XVL+RDY8WLyK6ZUYBhoPsXLesvzvGwYKr1NFpOnot7Qc47vu8j3HXrsJWRJ6NTqVbRt6cRJGq+lVNsqC3nP/lkAXwbwqpntr2x7ELNJ/iMzuwvAMQC3N6aLIlIPwWR39xeQvA7BjfXtjog0yuJ7LSIiVVGyi0RCyS4SCSW7SCSU7CKR0BDXOphewYeRLs3yoZxHi3wq6g3dvB79xvDqxFg6zadrLjv/f5823j6X4UM5h8k02Ju7hmjbUxNLaXy6xP9806nkawyKG3id3UJ19kVIZ3aRSCjZRSKhZBeJhJJdJBJKdpFIKNlFIqFkF4mE6uwXBJZsZiY28lrzGBlvDoRX/12XP0/jL564NDEWmoY6ZEPXWRo/PsLH4heLyVMub8rxOvtrOT5F93iBj2dny0UXlvG2wTmVavh7aRWd2UUioWQXiYSSXSQSSnaRSCjZRSKhZBeJhJJdJBKqs9cDH/KNsQKv2nbm+fLAw6XkMeEAr2WHxpv35Ydp/MrO4zT+H+XNNJ7J8HnrmXQHP7DFGX6uyqeTnzspwS9IcMnm2nbfEDqzi0RCyS4SCSW7SCSU7CKRULKLRELJLhIJJbtIJBayPvt6AN8DsBaz5cMd7v4tM3sYwJ8AuDAo+UF339WojrazjgL/n1ksB+rBgVr4q+fW0biT/U8VktdHB4DuFK/xTzkf9z083Enj2XzyePpj03zu9tCc9eXAcaX7nuTHPMRnqr9+oFUWclFNCcDX3P0VM+sB8LKZ7a7EHnX3v21c90SkXhayPvspAKcqt0fN7DCAixvdMRGprw/1OsjMLgVwFYC9lU33mtkBM3vczFYktNluZvvMbF8R/CWjiDTOgpPdzLoBPAngfncfAfBtAJsBbMPsmf/r87Vz9x3u3u/u/ZnwzF4i0iALSnYzy2A20b/v7j8GAHc/7e4z7l4G8B0A1zSumyJSq2Cym5kBeAzAYXf/xpztc6f+/AKAg/XvnojUy0I+jf8sgC8DeNXM9le2PQjgDjPbhtly3FEAdzekh4vA8s18uuX1PXwq6IkSL29d1v0uj/ecSYwtTU/Stv1dR2h8SyZ53wCwa+OVNH7V8uQhsg+tPkTb3lvoofFV3eM03sEGmk4vvtJZrRbyafwLAOabJDvKmrrIYqUr6EQioWQXiYSSXSQSSnaRSCjZRSKhZBeJhKaSvqCGIYtj+1fS+Esrl9N4boj/Gt6e3kTj+XeT68kWeFr/1ncdjU9dxHfQu5+fL47lkqea/sf1v0nbhhZFTk0E7nHlaGLosmODtGlwAOwiHOKqM7tIJJTsIpFQsotEQskuEgklu0gklOwikVCyi0TC3Ju3uKyZDQE4NmfTKgB8sHbrtGvf2rVfgPpWrXr2baO7r54v0NRk/8CDm+1z9/6WdYBo1761a78A9a1azeqbXsaLRELJLhKJVif7jhY/PtOufWvXfgHqW7Wa0reWvmcXkeZp9ZldRJpEyS4SiZYku5ndbGb/a2ZvmtkDrehDEjM7amavmtl+M9vX4r48bmaDZnZwzrZeM9ttZm9Uvs+7xl6L+vawmZ2sHLv9ZnZLi/q23syeM7NDZvaamd1X2d7SY0f61ZTj1vT37GaWAvA6gN8GcALASwDucHe+YkCTmNlRAP3u3vILMMzsNwCMAfieu19R2fY3AM66+yOVf5Qr3P3P2qRvDwMYa/Uy3pXVivrmLjMO4DYAf4wWHjvSr9vRhOPWijP7NQDedPcj7l4A8EMAt7agH23P3Z8H8P7lZm4FsLNyeydm/1iaLqFvbcHdT7n7K5XbowAuLDPe0mNH+tUUrUj2iwHMXRPoBNprvXcH8KyZvWxm21vdmXmsdfdTldsDANa2sjPzCC7j3UzvW2a8bY5dNcuf10of0H3Q9e5+NYDPAbin8nK1Lfnse7B2qp0uaBnvZplnmfFfauWxq3b581q1ItlPAlg/5+dLKtvagrufrHwfBPAU2m8p6tMXVtCtfOczJzZROy3jPd8y42iDY9fK5c9bkewvAdhiZpvMLAvgSwCeaUE/PsDMuiofnMDMugDchPZbivoZAHdWbt8J4OkW9uU92mUZ76RlxtHiY9fy5c/dvelfAG7B7CfybwH481b0IaFflwH4n8rXa63uG4AnMPuyrojZzzbuArASwB4AbwD4GYDeNurbPwB4FcABzCZWX4v6dj1mX6IfALC/8nVLq48d6VdTjpsulxWJhD6gE4mEkl0kEkp2kUgo2UUioWQXiYSSXSQSSnaRSPw/q4AZ2qtQjDEAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rFQyGMGthlsl"
      },
      "source": [
        "Finally, we define the corresponding item names for each class."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TUBbXxIzhjBt"
      },
      "source": [
        "# Define the class names.\n",
        "class_names = ['T-Shirt/Top', 'Trouser', 'Pullover', 'Dress', 'Coat',\n",
        "               'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle Boot']"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "J5ubfhUsFLR8"
      },
      "source": [
        "## 3. Neural Network\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-gPoq2dRFtrH"
      },
      "source": [
        "### 3.1. Define the Model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jU5sDVjhOVgF"
      },
      "source": [
        "We then have to define our neural network. Here, we define a sequential model with two fully-connected layers."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ri1rSpp3FSDS"
      },
      "source": [
        "# Define the sequential model.\n",
        "model = tf.keras.models.Sequential()"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5CNJ9gHWFWQv"
      },
      "source": [
        "# Define the network's layes.\n",
        "model.add(tf.keras.layers.Flatten(input_shape=(28, 28)))\n",
        "model.add(tf.keras.layers.Dense(128, activation='relu'))\n",
        "model.add(tf.keras.layers.Dense(10, activation='softmax'))"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cdMaLgtqFeV6",
        "outputId": "9ab3fc14-6230-4fa3-db72-1c0de6e01ea0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "# Show the model.\n",
        "model.summary()"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "flatten (Flatten)            (None, 784)               0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 128)               100480    \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 10)                1290      \n",
            "=================================================================\n",
            "Total params: 101,770\n",
            "Trainable params: 101,770\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GYactRmuOhCW"
      },
      "source": [
        "Once our model is defined, we can compile it using the Adam optimizer and the categorical cross-entropy loss function."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "57j-R0L4Fmmf"
      },
      "source": [
        "# Compile the model.\n",
        "model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hRVabAG8F3IL"
      },
      "source": [
        "### 3.2. Train the Model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0IuRzsr2Osga"
      },
      "source": [
        "We then train the model on 20 epochs, using a batch size of 256."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Yl1f9mvZF5Pc",
        "outputId": "ffe74c17-92d8-4f67-ee54-511d19e9ade3",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "# Define the parameters.\n",
        "num_epochs = 20\n",
        "batch_size = 256\n",
        "\n",
        "# Train the model.\n",
        "history = model.fit(X_train, \n",
        "                      y_train, \n",
        "                      epochs=num_epochs, \n",
        "                      batch_size=batch_size, \n",
        "                      validation_data=(X_test, y_test))"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/20\n",
            "235/235 [==============================] - 1s 6ms/step - loss: 0.6072 - accuracy: 0.7948 - val_loss: 0.5025 - val_accuracy: 0.8260\n",
            "Epoch 2/20\n",
            "235/235 [==============================] - 1s 5ms/step - loss: 0.4238 - accuracy: 0.8528 - val_loss: 0.4457 - val_accuracy: 0.8421\n",
            "Epoch 3/20\n",
            "235/235 [==============================] - 1s 5ms/step - loss: 0.3838 - accuracy: 0.8662 - val_loss: 0.4031 - val_accuracy: 0.8596\n",
            "Epoch 4/20\n",
            "235/235 [==============================] - 1s 5ms/step - loss: 0.3602 - accuracy: 0.8724 - val_loss: 0.3894 - val_accuracy: 0.8611\n",
            "Epoch 5/20\n",
            "235/235 [==============================] - 1s 5ms/step - loss: 0.3435 - accuracy: 0.8773 - val_loss: 0.3842 - val_accuracy: 0.8641\n",
            "Epoch 6/20\n",
            "235/235 [==============================] - 1s 6ms/step - loss: 0.3283 - accuracy: 0.8831 - val_loss: 0.3707 - val_accuracy: 0.8698\n",
            "Epoch 7/20\n",
            "235/235 [==============================] - 1s 6ms/step - loss: 0.3131 - accuracy: 0.8869 - val_loss: 0.3700 - val_accuracy: 0.8685\n",
            "Epoch 8/20\n",
            "235/235 [==============================] - 1s 6ms/step - loss: 0.2983 - accuracy: 0.8923 - val_loss: 0.3590 - val_accuracy: 0.8705\n",
            "Epoch 9/20\n",
            "235/235 [==============================] - 1s 6ms/step - loss: 0.2930 - accuracy: 0.8944 - val_loss: 0.3710 - val_accuracy: 0.8692\n",
            "Epoch 10/20\n",
            "235/235 [==============================] - 1s 5ms/step - loss: 0.2839 - accuracy: 0.8965 - val_loss: 0.3467 - val_accuracy: 0.8750\n",
            "Epoch 11/20\n",
            "235/235 [==============================] - 1s 6ms/step - loss: 0.2738 - accuracy: 0.9010 - val_loss: 0.3459 - val_accuracy: 0.8774\n",
            "Epoch 12/20\n",
            "235/235 [==============================] - 1s 5ms/step - loss: 0.2663 - accuracy: 0.9039 - val_loss: 0.3310 - val_accuracy: 0.8808\n",
            "Epoch 13/20\n",
            "235/235 [==============================] - 1s 5ms/step - loss: 0.2619 - accuracy: 0.9048 - val_loss: 0.3343 - val_accuracy: 0.8820\n",
            "Epoch 14/20\n",
            "235/235 [==============================] - 1s 6ms/step - loss: 0.2534 - accuracy: 0.9085 - val_loss: 0.3355 - val_accuracy: 0.8820\n",
            "Epoch 15/20\n",
            "235/235 [==============================] - 1s 5ms/step - loss: 0.2483 - accuracy: 0.9102 - val_loss: 0.3391 - val_accuracy: 0.8796\n",
            "Epoch 16/20\n",
            "235/235 [==============================] - 1s 5ms/step - loss: 0.2433 - accuracy: 0.9112 - val_loss: 0.3331 - val_accuracy: 0.8795\n",
            "Epoch 17/20\n",
            "235/235 [==============================] - 1s 5ms/step - loss: 0.2377 - accuracy: 0.9133 - val_loss: 0.3255 - val_accuracy: 0.8859\n",
            "Epoch 18/20\n",
            "235/235 [==============================] - 1s 5ms/step - loss: 0.2326 - accuracy: 0.9156 - val_loss: 0.3522 - val_accuracy: 0.8738\n",
            "Epoch 19/20\n",
            "235/235 [==============================] - 1s 5ms/step - loss: 0.2294 - accuracy: 0.9162 - val_loss: 0.3298 - val_accuracy: 0.8852\n",
            "Epoch 20/20\n",
            "235/235 [==============================] - 1s 5ms/step - loss: 0.2224 - accuracy: 0.9196 - val_loss: 0.3324 - val_accuracy: 0.8828\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WhyG0-sPHOAv"
      },
      "source": [
        "### 3.4. Display the Metrics"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1TFadfX_Oyg_"
      },
      "source": [
        "Finally, we display the metrics. We begin by displaying the model's accuracy and loss based on the test set."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vFoIAPdSYXqh",
        "outputId": "781d4844-3f2c-491e-c703-ba6f52458644",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "test_loss, test_accuracy = model.evaluate(X_test, y_test)\n",
        "print('Test Accuracy:', test_accuracy)\n",
        "print('Test Loss:', test_loss)"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "313/313 [==============================] - 0s 1ms/step - loss: 0.3324 - accuracy: 0.8828\n",
            "Test Accuracy: 0.8827999830245972\n",
            "Test Loss: 0.3323984742164612\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "k3p2c0-RO2M8"
      },
      "source": [
        "Then, we save the metric values for each epoch to plot the loss and accuracy curves for our model."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8-OdhC_0KtUx"
      },
      "source": [
        "# Save the metrics.\n",
        "metrics = history.history"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZztnSHgjZPDN"
      },
      "source": [
        "Finally, once we have our metric history, we can plot the curves."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yfosMMiqHQB0"
      },
      "source": [
        "# Save the loss values.\n",
        "training_loss_list = metrics['loss']\n",
        "test_loss_list = metrics['val_loss']"
      ],
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HKAZlyOmHV6f",
        "outputId": "52e27a89-7f52-436c-be95-135200ba2737",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "# Plot the training and test loss.\n",
        "x = np.arange(0, num_epochs, 1)\n",
        "plt.title('Training and Test Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Loss')\n",
        "plt.plot(x, training_loss_list, label='Training Loss')\n",
        "plt.plot(x, test_loss_list, label='Test Loss')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 16,
      "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": "RwzIUzuuI0ZP"
      },
      "source": [
        "train_accuracy_list = metrics['accuracy']\n",
        "test_accuracy_list = metrics['val_accuracy']"
      ],
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kHj-j2HpI2Lb",
        "outputId": "ebc37d8f-4354-49e9-e262-9ab70198e2d6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "plt.title('Training and Test Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.plot(x, train_accuracy_list, label='Training Accuracy')\n",
        "plt.plot(x, test_accuracy_list, label='Test Accuracy')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 18,
      "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": "markdown",
      "metadata": {
        "id": "x2BDwaUigDgC"
      },
      "source": [
        "## 4. Make a Prediction\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vx34e_AWgLeE"
      },
      "source": [
        "Once our model is trained, we can use it to make predictions. To do this, we first use our test set to predict the classes."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SylCmq61gTnM"
      },
      "source": [
        "# Make predictions with the trained model.\n",
        "predictions = model.predict(X_test)"
      ],
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uMo9sEZpggPq"
      },
      "source": [
        "Finally, we can show a random test image with its corresponding prediction."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dDn-I6xNiYqy",
        "outputId": "5ab548c2-46ca-4469-9ccc-22150e5ef884",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        }
      },
      "source": [
        "# Choose an index.\n",
        "index = 75\n",
        "\n",
        "# Show an image from the test set.\n",
        "plt.imshow(test_images[index])\n",
        "plt.show()"
      ],
      "execution_count": 20,
      "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": "ITg9K1criYq1",
        "outputId": "a2dcd959-83d5-47b1-afb1-166f7770dad9",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "print(\"Prediction:\", class_names[np.argmax(predictions[index])])"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Prediction: Dress\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}