Digit Recognizer/digit_recognizer.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.8"
    },
    "colab": {
      "name": "digit_recognizer.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cl_x1a4orrPV",
        "colab_type": "text"
      },
      "source": [
        "### Digit Recognizer\n",
        "\n",
        "---\n",
        "Kaggle competition | Link: https://www.kaggle.com/c/digit-recognizer/overview"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xX5st0fGrrPX",
        "colab_type": "text"
      },
      "source": [
        "#### Import necessary modules"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9wEdb8YOrrPY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from tensorflow.keras import models, layers"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "islw2YicrrPd",
        "colab_type": "text"
      },
      "source": [
        "#### Load and prepare the data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uNUmf1qWrrPe",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train = pd.read_csv('/content/train.csv')\n",
        "test = pd.read_csv('/content/test.csv')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DjOOuV6jrrPi",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 253
        },
        "outputId": "0ad5c51f-7d1b-4283-9647-d8844bf5b1ea"
      },
      "source": [
        "train.head()"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>label</th>\n",
              "      <th>pixel0</th>\n",
              "      <th>pixel1</th>\n",
              "      <th>pixel2</th>\n",
              "      <th>pixel3</th>\n",
              "      <th>pixel4</th>\n",
              "      <th>pixel5</th>\n",
              "      <th>pixel6</th>\n",
              "      <th>pixel7</th>\n",
              "      <th>pixel8</th>\n",
              "      <th>pixel9</th>\n",
              "      <th>pixel10</th>\n",
              "      <th>pixel11</th>\n",
              "      <th>pixel12</th>\n",
              "      <th>pixel13</th>\n",
              "      <th>pixel14</th>\n",
              "      <th>pixel15</th>\n",
              "      <th>pixel16</th>\n",
              "      <th>pixel17</th>\n",
              "      <th>pixel18</th>\n",
              "      <th>pixel19</th>\n",
              "      <th>pixel20</th>\n",
              "      <th>pixel21</th>\n",
              "      <th>pixel22</th>\n",
              "      <th>pixel23</th>\n",
              "      <th>pixel24</th>\n",
              "      <th>pixel25</th>\n",
              "      <th>pixel26</th>\n",
              "      <th>pixel27</th>\n",
              "      <th>pixel28</th>\n",
              "      <th>pixel29</th>\n",
              "      <th>pixel30</th>\n",
              "      <th>pixel31</th>\n",
              "      <th>pixel32</th>\n",
              "      <th>pixel33</th>\n",
              "      <th>pixel34</th>\n",
              "      <th>pixel35</th>\n",
              "      <th>pixel36</th>\n",
              "      <th>pixel37</th>\n",
              "      <th>pixel38</th>\n",
              "      <th>...</th>\n",
              "      <th>pixel744</th>\n",
              "      <th>pixel745</th>\n",
              "      <th>pixel746</th>\n",
              "      <th>pixel747</th>\n",
              "      <th>pixel748</th>\n",
              "      <th>pixel749</th>\n",
              "      <th>pixel750</th>\n",
              "      <th>pixel751</th>\n",
              "      <th>pixel752</th>\n",
              "      <th>pixel753</th>\n",
              "      <th>pixel754</th>\n",
              "      <th>pixel755</th>\n",
              "      <th>pixel756</th>\n",
              "      <th>pixel757</th>\n",
              "      <th>pixel758</th>\n",
              "      <th>pixel759</th>\n",
              "      <th>pixel760</th>\n",
              "      <th>pixel761</th>\n",
              "      <th>pixel762</th>\n",
              "      <th>pixel763</th>\n",
              "      <th>pixel764</th>\n",
              "      <th>pixel765</th>\n",
              "      <th>pixel766</th>\n",
              "      <th>pixel767</th>\n",
              "      <th>pixel768</th>\n",
              "      <th>pixel769</th>\n",
              "      <th>pixel770</th>\n",
              "      <th>pixel771</th>\n",
              "      <th>pixel772</th>\n",
              "      <th>pixel773</th>\n",
              "      <th>pixel774</th>\n",
              "      <th>pixel775</th>\n",
              "      <th>pixel776</th>\n",
              "      <th>pixel777</th>\n",
              "      <th>pixel778</th>\n",
              "      <th>pixel779</th>\n",
              "      <th>pixel780</th>\n",
              "      <th>pixel781</th>\n",
              "      <th>pixel782</th>\n",
              "      <th>pixel783</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5 rows × 785 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "   label  pixel0  pixel1  pixel2  ...  pixel780  pixel781  pixel782  pixel783\n",
              "0      1       0       0       0  ...         0         0         0         0\n",
              "1      0       0       0       0  ...         0         0         0         0\n",
              "2      1       0       0       0  ...         0         0         0         0\n",
              "3      4       0       0       0  ...         0         0         0         0\n",
              "4      0       0       0       0  ...         0         0         0         0\n",
              "\n",
              "[5 rows x 785 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UIW5jaR4rrPn",
        "colab_type": "text"
      },
      "source": [
        "As we can see, apart from **label**, dataset has $784$ more columns, which represnets pixle values of an $28$ x $28$ image.\n",
        "\n",
        "Let's create `X_train` and `y_train` from `train`."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iiVZ8ctBrrPp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_train = train['label']\n",
        "X_train = train.drop(['label'], axis=1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mCNCSHQQrrPt",
        "colab_type": "text"
      },
      "source": [
        "It's time for some normalization and reshaping."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T5v-fTberrPu",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X_train /= 255.0\n",
        "test /= 255.0"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HkP-QW_PrrPy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X_train = X_train.values.reshape(-1, 28, 28, 1)\n",
        "test = test.values.reshape(-1, 28, 28, 1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5OjrLpFJrrP2",
        "colab_type": "text"
      },
      "source": [
        "Now, as we are ready with our data, it's time for defining and fitting a model. But before that, let's visualize what we have."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mmfEzwyfrrP3",
        "colab_type": "text"
      },
      "source": [
        "#### Visualization"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "anwOKyBorrP4",
        "colab_type": "text"
      },
      "source": [
        "Following function plots $12$ images from the dataset in a $3$ x $4$ grid. Title of each image represents its corresponding label."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cbSaPg9QrrP5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def plot(data, labels, title='Label'):\n",
        "    plt.figure(figsize=(10, 9))\n",
        "    for i in range(12):\n",
        "        plt.subplot(3, 4, i+1)\n",
        "        plt.imshow(data[i][:,:,0])\n",
        "        plt.title('{}: {}'.format(title, labels[i]))\n",
        "        plt.axis('off');"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7fT9wGUzrrP_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 531
        },
        "outputId": "6f0b7d08-da07-489a-ac3d-c1a0ee5a5610"
      },
      "source": [
        "plot(X_train, y_train)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlAAAAICCAYAAAAApPO5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRAIFAADgiQQKAADA0/8D+QxRgJZwV+QAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 720x648 with 12 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LUM_ZdQwrrQF",
        "colab_type": "text"
      },
      "source": [
        "#### Define the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mdkCRhhQrrQG",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 105
        },
        "outputId": "b4e13a18-bf75-469c-e90b-1b0fc5657b3f"
      },
      "source": [
        "model = models.Sequential([\n",
        "    layers.Conv2D(32, (3, 3), activation='relu', padding = 'Same', input_shape=(28, 28, 1)),\n",
        "    layers.MaxPooling2D(),\n",
        "    layers.Conv2D(64, (3, 3), activation='relu', padding = 'Same'),\n",
        "    layers.MaxPooling2D(),\n",
        "    layers.Conv2D(128, (3, 3), activation='relu'),\n",
        "    layers.MaxPooling2D(),\n",
        "    layers.Flatten(),\n",
        "    layers.Dense(512, activation='relu'),\n",
        "    layers.Dense(10, activation='softmax')\n",
        "])"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING: Logging before flag parsing goes to stderr.\n",
            "W0901 16:42:33.077652 140621963966336 deprecation.py:506] From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/init_ops.py:1251: calling VarianceScaling.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Call initializer instance with the dtype argument instead of passing it to the constructor\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mAXk354vrrQK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.compile(optimizer='adam',\n",
        "             loss='sparse_categorical_crossentropy',\n",
        "             metrics=['acc'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "K_7asJtMrrQN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 459
        },
        "outputId": "fafd9a97-935c-4b26-b555-1b641a303c19"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 28, 28, 32)        320       \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 14, 14, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 14, 14, 64)        18496     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 7, 7, 64)          0         \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 5, 5, 128)         73856     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 2, 2, 128)         0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 512)               0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 512)               262656    \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 10)                5130      \n",
            "=================================================================\n",
            "Total params: 360,458\n",
            "Trainable params: 360,458\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "O_nIAAuIrrQR",
        "colab_type": "text"
      },
      "source": [
        "#### Fit the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bc3abG1xrrQS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "EPOCHS = 5"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ICvbe-ZlrrQW",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "outputId": "16052968-9087-4aa7-a85e-43dcaf69c876"
      },
      "source": [
        "%%time\n",
        "history = model.fit(X_train, y_train.values,\n",
        "                   validation_split=.1,\n",
        "                   epochs=EPOCHS, batch_size=64,\n",
        "                   verbose=2)"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 37800 samples, validate on 4200 samples\n",
            "Epoch 1/5\n",
            "37800/37800 - 50s - loss: 0.2101 - acc: 0.9348 - val_loss: 0.0523 - val_acc: 0.9845\n",
            "Epoch 2/5\n",
            "37800/37800 - 49s - loss: 0.0521 - acc: 0.9838 - val_loss: 0.0582 - val_acc: 0.9826\n",
            "Epoch 3/5\n",
            "37800/37800 - 49s - loss: 0.0366 - acc: 0.9887 - val_loss: 0.0379 - val_acc: 0.9886\n",
            "Epoch 4/5\n",
            "37800/37800 - 49s - loss: 0.0282 - acc: 0.9911 - val_loss: 0.0556 - val_acc: 0.9833\n",
            "Epoch 5/5\n",
            "37800/37800 - 49s - loss: 0.0214 - acc: 0.9926 - val_loss: 0.0224 - val_acc: 0.9924\n",
            "CPU times: user 7min 48s, sys: 7.16 s, total: 7min 55s\n",
            "Wall time: 4min 8s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "V9KHhPwbrrQa",
        "colab_type": "text"
      },
      "source": [
        "Plot Accuracy vs. Loss!"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C1XZSKHerrQb",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "acc = history.history['acc']\n",
        "val_acc = history.history['val_acc']\n",
        "\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0O9rSqs2rrQe",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "epochs = range(EPOCHS)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HlzdrB2qrrQi",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 404
        },
        "outputId": "83fb1e7c-99fb-44f5-ab7a-790a8d210291"
      },
      "source": [
        "plt.figure(figsize=(15, 6))\n",
        "\n",
        "plt.subplot(1, 2, 1)\n",
        "plt.title('Training and Validation Loss')\n",
        "plt.plot(epochs, loss, label='Training')\n",
        "plt.plot(epochs, val_loss, label='Validation')\n",
        "plt.legend()\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Loss')\n",
        "\n",
        "plt.subplot(1, 2, 2)\n",
        "plt.title('Training and Validation Accuracy')\n",
        "plt.plot(epochs, acc, label='Training')\n",
        "plt.plot(epochs, val_acc, label='Validation')\n",
        "plt.legend()\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Accuracy')\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA4gAAAGDCAYAAABp6D4kAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 1080x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e8tIadIlrrQm",
        "colab_type": "text"
      },
      "source": [
        "#### Make predictions"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Lr9tW36frrQn",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "8944ab08-f52a-47e9-bd89-533a7179e099"
      },
      "source": [
        "%%time\n",
        "results = model.predict(test)"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "CPU times: user 19.7 s, sys: 331 ms, total: 20 s\n",
            "Wall time: 10.7 s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_Bho1MWxrrQu",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "results = np.argmax(results, axis=1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FwG7hvz-rrQ1",
        "colab_type": "text"
      },
      "source": [
        "Let's see some of our predictions."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WtfDab_MrrQ4",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 531
        },
        "outputId": "3a1b4329-bb9e-49b7-8c44-85ffeb4783e2"
      },
      "source": [
        "plot(test, results, 'Predicted label')"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlAAAAICCAYAAAAApPO5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 720x648 with 12 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HtN9C-0LrrQ-",
        "colab_type": "text"
      },
      "source": [
        "#### Export predictions"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0ytGCb7grrQ_",
        "colab_type": "text"
      },
      "source": [
        "Convert `results` to **Pandas Series**."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c-qH_CfIrrRD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "results = pd.Series(results, name='Label')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YDCkASVarrRH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "submission = pd.concat([pd.Series(range(1, 28001), name='ImageId'), results], axis=1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sT3WeYk6rrRL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "submission.to_csv('digit_recognizer.csv', index=False)"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}