MNIST Keras Tutorial.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Library imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\", category=FutureWarning)\n",
    "from keras.utils import to_categorical\n",
    "import numpy as np\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense,Conv2D,MaxPooling2D,Flatten, Dropout\n",
    "from matplotlib import pyplot as plt\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "traindf = pd.read_csv(\"Data/train.csv\")\n",
    "testdf = pd.read_csv(\"Data/test.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(42000, 785)\n"
     ]
    },
    {
     "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>...</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>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <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>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <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>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <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>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <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>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <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  pixel3  pixel4  pixel5  pixel6  pixel7  \\\n",
       "0      1       0       0       0       0       0       0       0       0   \n",
       "1      0       0       0       0       0       0       0       0       0   \n",
       "2      1       0       0       0       0       0       0       0       0   \n",
       "3      4       0       0       0       0       0       0       0       0   \n",
       "4      0       0       0       0       0       0       0       0       0   \n",
       "\n",
       "   pixel8    ...     pixel774  pixel775  pixel776  pixel777  pixel778  \\\n",
       "0       0    ...            0         0         0         0         0   \n",
       "1       0    ...            0         0         0         0         0   \n",
       "2       0    ...            0         0         0         0         0   \n",
       "3       0    ...            0         0         0         0         0   \n",
       "4       0    ...            0         0         0         0         0   \n",
       "\n",
       "   pixel779  pixel780  pixel781  pixel782  pixel783  \n",
       "0         0         0         0         0         0  \n",
       "1         0         0         0         0         0  \n",
       "2         0         0         0         0         0  \n",
       "3         0         0         0         0         0  \n",
       "4         0         0         0         0         0  \n",
       "\n",
       "[5 rows x 785 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(traindf.shape)\n",
    "traindf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(28000, 784)\n"
     ]
    },
    {
     "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>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>...</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>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>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",
       "    </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>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <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>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</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>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <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>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <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 × 784 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   pixel0  pixel1  pixel2  pixel3  pixel4  pixel5  pixel6  pixel7  pixel8  \\\n",
       "0       0       0       0       0       0       0       0       0       0   \n",
       "1       0       0       0       0       0       0       0       0       0   \n",
       "2       0       0       0       0       0       0       0       0       0   \n",
       "3       0       0       0       0       0       0       0       0       0   \n",
       "4       0       0       0       0       0       0       0       0       0   \n",
       "\n",
       "   pixel9    ...     pixel774  pixel775  pixel776  pixel777  pixel778  \\\n",
       "0       0    ...            0         0         0         0         0   \n",
       "1       0    ...            0         0         0         0         0   \n",
       "2       0    ...            0         0         0         0         0   \n",
       "3       0    ...            0         0         0         0         0   \n",
       "4       0    ...            0         0         0         0         0   \n",
       "\n",
       "   pixel779  pixel780  pixel781  pixel782  pixel783  \n",
       "0         0         0         0         0         0  \n",
       "1         0         0         0         0         0  \n",
       "2         0         0         0         0         0  \n",
       "3         0         0         0         0         0  \n",
       "4         0         0         0         0         0  \n",
       "\n",
       "[5 rows x 784 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(testdf.shape)\n",
    "testdf.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are 10 classes to classify the images into (0-9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "42000\n",
      "10\n"
     ]
    }
   ],
   "source": [
    "train_lbls = traindf.label\n",
    "print(len(train_lbls))\n",
    "print(len(np.unique(train_lbls)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Shape of the test and train labels to see how many test and train images we have"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(42000, 784)\n"
     ]
    }
   ],
   "source": [
    "X_train = traindf.drop(labels = [\"label\"],axis = 1)\n",
    "print(X_train.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have 42,000 images, each one is 28x28 "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Converting values to lie between 0 and 1 (since they are currently digits between 0 and 255, denoting how black each pixel is)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = X_train / 255.0\n",
    "test = testdf.as_matrix() / 255.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_data = test.astype('float32')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Reshaping the training and test data so the neural network can take it as an input (4 dimensions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = X_train.values.reshape(-1,28,28,1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_data = test.reshape(-1, 28, 28, 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Keras uses one hot encoding\n",
    "\n",
    "Hence changing the labels from integer to categorical data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original label :  1\n",
      "After conversion to categorical ( one-hot ) :  [0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n"
     ]
    }
   ],
   "source": [
    "train_labels_one_hot = to_categorical(train_lbls)\n",
    "\n",
    "#Display the category label using one-hot encoding\n",
    "print('Original label : ', train_lbls[2])\n",
    "print('After conversion to categorical ( one-hot ) : ', train_labels_one_hot[2])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Our data is now ready to be inputted into a neural network**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Making the Keras Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "model=Sequential()\n",
    "\n",
    "model.add(Conv2D(kernel_size=5,strides=1,filters=64,\n",
    "                padding='same',activation='relu', input_shape = (28,28,1)))\n",
    "model.add(Conv2D(kernel_size=5,strides=1,filters=64,\n",
    "                padding='same',activation='relu'))\n",
    "\n",
    "model.add(MaxPooling2D(pool_size=2,strides=2))\n",
    "model.add(Dropout(0.25))\n",
    "\n",
    "model.add(Conv2D(kernel_size=5,strides=1,filters=64,\n",
    "                padding='same',activation='relu'))\n",
    "model.add(Conv2D(kernel_size=5,strides=1,filters=64,\n",
    "                padding='same',activation='relu'))\n",
    "\n",
    "model.add(MaxPooling2D(pool_size=2,strides=2))\n",
    "model.add(Dropout(0.25))\n",
    "\n",
    "model.add(Flatten())\n",
    "model.add(Dense(128,activation='relu'))\n",
    "model.add(Dense(10,activation='softmax'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Adding dropout layers to reduce overfitting\n",
    "\n",
    "[Read more about layers here](https://forums.fast.ai/t/dense-vs-convolutional-vs-fully-connected-layers/191/2 \"Layers info.\")\n",
    "\n",
    "The last Dense layer has 10 units (since we need to classify into 10 classes (0-9))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_7 (Conv2D)            (None, 28, 28, 64)        1664      \n",
      "_________________________________________________________________\n",
      "conv2d_8 (Conv2D)            (None, 28, 28, 64)        102464    \n",
      "_________________________________________________________________\n",
      "max_pooling2d_4 (MaxPooling2 (None, 14, 14, 64)        0         \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 14, 14, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_9 (Conv2D)            (None, 14, 14, 64)        102464    \n",
      "_________________________________________________________________\n",
      "conv2d_10 (Conv2D)           (None, 14, 14, 64)        102464    \n",
      "_________________________________________________________________\n",
      "max_pooling2d_5 (MaxPooling2 (None, 7, 7, 64)          0         \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (None, 7, 7, 64)          0         \n",
      "_________________________________________________________________\n",
      "flatten_2 (Flatten)          (None, 3136)              0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 128)               401536    \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 10)                1290      \n",
      "=================================================================\n",
      "Total params: 711,882\n",
      "Trainable params: 711,882\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using the Adam optimizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.optimizers import Adam\n",
    "optimizer=Adam(lr=1e-3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Compiling the model, using the Adam optimizer, with a categorical crossentropy loss so we can predict into multiple classes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer=optimizer,loss='categorical_crossentropy',metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fitting out data onto the model, with 256 images at a time, for 10 epochs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "42000/42000 [==============================] - 93s 2ms/step - loss: 0.2688 - acc: 0.9137\n",
      "Epoch 2/10\n",
      "42000/42000 [==============================] - 83s 2ms/step - loss: 0.0556 - acc: 0.9828\n",
      "Epoch 3/10\n",
      "42000/42000 [==============================] - 82s 2ms/step - loss: 0.0401 - acc: 0.9874\n",
      "Epoch 4/10\n",
      "42000/42000 [==============================] - 82s 2ms/step - loss: 0.0311 - acc: 0.9900\n",
      "Epoch 5/10\n",
      "42000/42000 [==============================] - 83s 2ms/step - loss: 0.0238 - acc: 0.9924\n",
      "Epoch 6/10\n",
      "42000/42000 [==============================] - 83s 2ms/step - loss: 0.0210 - acc: 0.9932\n",
      "Epoch 7/10\n",
      "42000/42000 [==============================] - 83s 2ms/step - loss: 0.0267 - acc: 0.9918\n",
      "Epoch 8/10\n",
      "42000/42000 [==============================] - 83s 2ms/step - loss: 0.0157 - acc: 0.9951\n",
      "Epoch 9/10\n",
      "42000/42000 [==============================] - 83s 2ms/step - loss: 0.0139 - acc: 0.9956\n",
      "Epoch 10/10\n",
      "42000/42000 [==============================] - 83s 2ms/step - loss: 0.0119 - acc: 0.9962\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(x=X_train, y=train_labels_one_hot,\n",
    "                    batch_size=256, epochs=10, verbose=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plotting the accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,1,'Accuracy Plot')"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x277ae49ca20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "#Checking for overfitting\n",
    "\n",
    "#Plot the Loss Curves\n",
    "plt.figure(figsize=[8,6])\n",
    "plt.plot(history.history['acc'],'r',linewidth=3.0)\n",
    "plt.xlabel('Epochs ',fontsize=16)\n",
    "plt.ylabel('Accuracy',fontsize=16)\n",
    "plt.title('Accuracy Plot',fontsize=16)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Predicting the values of the test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "28000"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions = model.predict(test_data)\n",
    "\n",
    "predictions = [np.argmax(i) for i in predictions]\n",
    "len(predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plotting some image predictions from our neural model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "img_shape=(28,28)\n",
    "\n",
    "def plot_images(images, cls_pred=None):\n",
    "    assert len(images) == 9\n",
    "    \n",
    "    # Create figure with 3x3 sub-plots.\n",
    "    fig, axes = plt.subplots(3, 3)\n",
    "    fig.subplots_adjust(hspace=0.3, wspace=0.3)\n",
    "\n",
    "    for i, ax in enumerate(axes.flat):\n",
    "        # Plot image.\n",
    "        ax.imshow(images[i].reshape(img_shape), cmap='binary')\n",
    "\n",
    "        # Show true and predicted classes.\n",
    "        xlabel = \"Pred: {0}\".format(cls_pred[i])\n",
    "\n",
    "        # Show the classes as the label on the x-axis.\n",
    "        ax.set_xlabel(xlabel)\n",
    "        \n",
    "        # Remove ticks from the plot.\n",
    "        ax.set_xticks([])\n",
    "        ax.set_yticks([])\n",
    "    \n",
    "    # Ensure the plot is shown correctly with multiple plots\n",
    "    # in a single Notebook cell.\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([3, 5, 7, 4, 0, 4, 3, 3, 1], dtype=int64)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "images = test_data[9:18]\n",
    "\n",
    "y_pred=model.predict(x=images)\n",
    "cls_pred=np.argmax(y_pred,axis=1)\n",
    "\n",
    "cls_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = model.predict(x=images)\n",
    "cls_pred = np.argmax(y_pred, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x27788263978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_images(images=images,\n",
    "           cls_pred=cls_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hence we have used Neural Networks to successfully predict handwritten image classes, with pretty high accuracy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Making a submission file for Kaggle:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "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>ImageId</th>\n",
       "      <th>Label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   ImageId  Label\n",
       "0        1      2\n",
       "1        2      0\n",
       "2        3      9\n",
       "3        4      0\n",
       "4        5      3"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = list(zip(list(range(1, 28001)), predictions))\n",
    "\n",
    "submission_df = pd.DataFrame(data, columns=['ImageId','Label'])\n",
    "submission_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission_df.to_csv(\"Submission.csv\",index=False)"
   ]
  }
 ],
 "metadata": {
  "hide_input": false,
  "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": "3.5.5"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}