model_architecture.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "bdc7b663",
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "from tensorflow.keras.layers import Conv2D, MaxPooling2D, BatchNormalization, Flatten, Dense, Dropout\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from PIL import Image\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.preprocessing import OneHotEncoder, LabelEncoder\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "b345f651",
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_model(inputs,output_size):\n",
    "    inputs = tf.keras.Input(shape=inputs.shape)\n",
    "    x = Conv2D(filters=64, kernel_size=3, activation = 'relu', padding='same')(inputs)\n",
    "    x  = MaxPooling2D(pool_size=2)(x)\n",
    "    x = BatchNormalization()(x)\n",
    "    x = Conv2D(filters=32, kernel_size=3, activation='relu', padding='same')(x)\n",
    "    x  = MaxPooling2D(pool_size=2)(x)\n",
    "    x = BatchNormalization()(x)\n",
    "    x = Conv2D(filters=16, kernel_size=3, activation='relu', padding='same')(x)\n",
    "    x = BatchNormalization()(x)\n",
    "    x = Flatten()(x)\n",
    "    x = Dense(512, activation='relu')(x)\n",
    "    x = Dense(256, activation='relu')(x)\n",
    "    x = Dropout(0.3)(x)\n",
    "    x = Dense(output_size, activation='softmax')(x)\n",
    "    \n",
    "    return inputs, x\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "55261545",
   "metadata": {},
   "outputs": [],
   "source": [
    "output_size = 62\n",
    "inputs = np.zeros((28, 28, 1))\n",
    "input_, output = create_model(inputs,output_size = 62)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "6b031dd0",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = tf.keras.Model(inputs=input_, outputs = output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "80a7a635",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_1\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_2 (InputLayer)        [(None, 28, 28, 1)]       0         \n",
      "                                                                 \n",
      " conv2d_3 (Conv2D)           (None, 28, 28, 64)        640       \n",
      "                                                                 \n",
      " max_pooling2d_2 (MaxPooling  (None, 14, 14, 64)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " batch_normalization_3 (Batc  (None, 14, 14, 64)       256       \n",
      " hNormalization)                                                 \n",
      "                                                                 \n",
      " conv2d_4 (Conv2D)           (None, 14, 14, 32)        18464     \n",
      "                                                                 \n",
      " max_pooling2d_3 (MaxPooling  (None, 7, 7, 32)         0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " batch_normalization_4 (Batc  (None, 7, 7, 32)         128       \n",
      " hNormalization)                                                 \n",
      "                                                                 \n",
      " conv2d_5 (Conv2D)           (None, 7, 7, 16)          4624      \n",
      "                                                                 \n",
      " batch_normalization_5 (Batc  (None, 7, 7, 16)         64        \n",
      " hNormalization)                                                 \n",
      "                                                                 \n",
      " flatten_1 (Flatten)         (None, 784)               0         \n",
      "                                                                 \n",
      " dense_3 (Dense)             (None, 512)               401920    \n",
      "                                                                 \n",
      " dense_4 (Dense)             (None, 256)               131328    \n",
      "                                                                 \n",
      " dropout_1 (Dropout)         (None, 256)               0         \n",
      "                                                                 \n",
      " dense_5 (Dense)             (None, 62)                15934     \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 573,358\n",
      "Trainable params: 573,134\n",
      "Non-trainable params: 224\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1ac333d",
   "metadata": {},
   "source": [
    "### Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "7abbacc7",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = pd.read_csv('dataset_final_2.csv', low_memory=False, header=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "c392c86e",
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>775</th>\n",
       "      <th>776</th>\n",
       "      <th>777</th>\n",
       "      <th>778</th>\n",
       "      <th>779</th>\n",
       "      <th>780</th>\n",
       "      <th>781</th>\n",
       "      <th>782</th>\n",
       "      <th>783</th>\n",
       "      <th>784</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>1</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>1</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>4</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",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>422513</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>Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>422514</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>Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>422515</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>Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>422516</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>Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>422517</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>Z</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>422518 rows × 785 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        0    1    2    3    4    5    6    7    8    9    ...  775  776  777  \\\n",
       "0         0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "1         0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "2         0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "3         0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "4         0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "...     ...  ...  ...  ...  ...  ...  ...  ...  ...  ...  ...  ...  ...  ...   \n",
       "422513    0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "422514    0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "422515    0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "422516    0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "422517    0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "\n",
       "        778  779  780  781  782  783  784  \n",
       "0         0    0    0    0    0    0    1  \n",
       "1         0    0    0    0    0    0    0  \n",
       "2         0    0    0    0    0    0    1  \n",
       "3         0    0    0    0    0    0    4  \n",
       "4         0    0    0    0    0    0    0  \n",
       "...     ...  ...  ...  ...  ...  ...  ...  \n",
       "422513    0    0    0    0    0    0    Z  \n",
       "422514    0    0    0    0    0    0    Z  \n",
       "422515    0    0    0    0    0    0    Z  \n",
       "422516    0    0    0    0    0    0    Z  \n",
       "422517    0    0    0    0    0    0    Z  \n",
       "\n",
       "[422518 rows x 785 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "e9b84bbb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['1', '0', '4', '7', '3', '5', '8', '9', '2', '6', 'A', 'B', 'C',\n",
       "       'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P',\n",
       "       'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', 'a', 'b', 'c',\n",
       "       'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p',\n",
       "       'q', 'r', 's', 'u', 'v', 'w', 'x', 'y', 'z', 't'], dtype=object)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset.iloc[:,784].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "fe2c5e76",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_shuffled = dataset.sample(frac=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "00bea78c",
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>775</th>\n",
       "      <th>776</th>\n",
       "      <th>777</th>\n",
       "      <th>778</th>\n",
       "      <th>779</th>\n",
       "      <th>780</th>\n",
       "      <th>781</th>\n",
       "      <th>782</th>\n",
       "      <th>783</th>\n",
       "      <th>784</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1892</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>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>391469</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>W</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>386983</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>V</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>348384</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>T</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>392657</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>W</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>225063</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>O</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>334397</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>T</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>313997</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>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>185971</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>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38286</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>422518 rows × 785 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        0    1    2    3    4    5    6    7    8    9    ...  775  776  777  \\\n",
       "1892      0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "391469    0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "386983    0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "348384    0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "392657    0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "...     ...  ...  ...  ...  ...  ...  ...  ...  ...  ...  ...  ...  ...  ...   \n",
       "225063    0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "334397    0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "313997    0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "185971    0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "38286     0    0    0    0    0    0    0    0    0    0  ...    0    0    0   \n",
       "\n",
       "        778  779  780  781  782  783  784  \n",
       "1892      0    0    0    0    0    0    9  \n",
       "391469    0    0    0    0    0    0    W  \n",
       "386983    0    0    0    0    0    0    V  \n",
       "348384    0    0    0    0    0    0    T  \n",
       "392657    0    0    0    0    0    0    W  \n",
       "...     ...  ...  ...  ...  ...  ...  ...  \n",
       "225063    0    0    0    0    0    0    O  \n",
       "334397    0    0    0    0    0    0    T  \n",
       "313997    0    0    0    0    0    0    S  \n",
       "185971    0    0    0    0    0    0    N  \n",
       "38286     0    0    0    0    0    0    0  \n",
       "\n",
       "[422518 rows x 785 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset_shuffled"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "320cee26",
   "metadata": {},
   "outputs": [],
   "source": [
    "x = dataset_shuffled.iloc[:,0:784].values\n",
    "y = dataset_shuffled.iloc[:,784:].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "3af4c5ac",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['J'], dtype=object)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "i=1784\n",
    "plt.imshow(x[i].reshape(28,28), cmap='gray')\n",
    "y[i]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "09cdce20",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23\n",
      " 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47\n",
      " 48 49 50 51 52 53 54 55 56 57 58 59 60 61]\n",
      "['J']\n"
     ]
    }
   ],
   "source": [
    "le = LabelEncoder()\n",
    "y_encoded = le.fit_transform(y.reshape(-1,))\n",
    "print(np.unique(y_encoded))\n",
    "print(le.inverse_transform([y_encoded[1784]]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "aecf2c87",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_encoded = y_encoded.reshape(-1,1)\n",
    "ct = ColumnTransformer(transformers=[('encoder', OneHotEncoder(), [0])], remainder='drop')\n",
    "y_ct = ct.fit_transform(y_encoded)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "c83d1801",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0. 0. 0. ... 0. 0. 0.]\n",
      " [0. 0. 0. ... 0. 0. 0.]\n",
      " [0. 0. 0. ... 0. 0. 0.]\n",
      " ...\n",
      " [0. 0. 0. ... 0. 0. 0.]\n",
      " [0. 0. 0. ... 0. 0. 0.]\n",
      " [1. 0. 0. ... 0. 0. 0.]]\n"
     ]
    }
   ],
   "source": [
    "y_ct = y_ct.toarray()\n",
    "print(y_ct)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "0c4b7cff",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_ct = y_ct.astype(np.float32)\n",
    "x_train, x_test, y_train, y_test = train_test_split(x, y_ct, test_size=0.4, random_state=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "ce470c73",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x_train shape: (253510, 784)\n",
      "x_test shape: (169008, 784)\n",
      "y_train shape: (253510, 62)\n",
      "y_test shape: (169008, 62)\n"
     ]
    }
   ],
   "source": [
    "print(f\"x_train shape: {x_train.shape}\")\n",
    "print(f\"x_test shape: {x_test.shape}\")\n",
    "print(f\"y_train shape: {y_train.shape}\")\n",
    "print(f\"y_test shape: {y_test.shape}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "a676ffc2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['S'], dtype=object)"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(x_train[10].reshape(28,28), cmap = 'gray')\n",
    "le.inverse_transform([np.argmax(y_train[10])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "f938dcd9",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_datagen = ImageDataGenerator(rescale=1./255)\n",
    "validation_datagen = ImageDataGenerator(rescale=1./255)\n",
    "\n",
    "train_datagen.fit(x_train.reshape(-1, 28, 28, 1))\n",
    "validation_datagen.fit(x_test.reshape(-1, 28, 28, 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "fd82a158",
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = x_train.reshape(-1, 28, 28, 1)\n",
    "x_test = x_test.reshape(-1, 28, 28, 1)\n",
    "\n",
    "model.compile(loss=tf.keras.losses.categorical_crossentropy,\n",
    "             optimizer=tf.keras.optimizers.Adam(learning_rate = 0.0001),\n",
    "             metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "a6c3edd7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "3962/3962 [==============================] - 53s 11ms/step - loss: 0.3948 - accuracy: 0.8963 - val_loss: 0.1533 - val_accuracy: 0.9579\n",
      "Epoch 2/20\n",
      "3962/3962 [==============================] - 42s 11ms/step - loss: 0.1476 - accuracy: 0.9585 - val_loss: 0.1111 - val_accuracy: 0.9695\n",
      "Epoch 3/20\n",
      "3962/3962 [==============================] - 41s 10ms/step - loss: 0.1083 - accuracy: 0.9690 - val_loss: 0.1021 - val_accuracy: 0.9709\n",
      "Epoch 4/20\n",
      "3962/3962 [==============================] - 42s 11ms/step - loss: 0.0868 - accuracy: 0.9749 - val_loss: 0.0839 - val_accuracy: 0.9763\n",
      "Epoch 5/20\n",
      "3962/3962 [==============================] - 41s 10ms/step - loss: 0.0724 - accuracy: 0.9785 - val_loss: 0.0772 - val_accuracy: 0.9778\n",
      "Epoch 6/20\n",
      "3962/3962 [==============================] - 41s 10ms/step - loss: 0.0607 - accuracy: 0.9814 - val_loss: 0.0722 - val_accuracy: 0.9793\n",
      "Epoch 7/20\n",
      "3962/3962 [==============================] - 41s 10ms/step - loss: 0.0528 - accuracy: 0.9838 - val_loss: 0.0697 - val_accuracy: 0.9804\n",
      "Epoch 8/20\n",
      "3962/3962 [==============================] - 41s 10ms/step - loss: 0.0450 - accuracy: 0.9860 - val_loss: 0.0662 - val_accuracy: 0.9818\n",
      "Epoch 9/20\n",
      "3962/3962 [==============================] - 337s 85ms/step - loss: 0.0394 - accuracy: 0.9876 - val_loss: 0.0698 - val_accuracy: 0.9806\n",
      "Epoch 10/20\n",
      "3962/3962 [==============================] - 36s 9ms/step - loss: 0.0348 - accuracy: 0.9889 - val_loss: 0.0677 - val_accuracy: 0.9818\n",
      "Epoch 11/20\n",
      "3962/3962 [==============================] - 36s 9ms/step - loss: 0.0310 - accuracy: 0.9901 - val_loss: 0.0640 - val_accuracy: 0.9831\n",
      "Epoch 12/20\n",
      "3962/3962 [==============================] - 36s 9ms/step - loss: 0.0270 - accuracy: 0.9913 - val_loss: 0.0645 - val_accuracy: 0.9829\n",
      "Epoch 13/20\n",
      "3962/3962 [==============================] - 37s 9ms/step - loss: 0.0244 - accuracy: 0.9919 - val_loss: 0.0692 - val_accuracy: 0.9825\n",
      "Epoch 14/20\n",
      "3962/3962 [==============================] - 36s 9ms/step - loss: 0.0223 - accuracy: 0.9926 - val_loss: 0.0663 - val_accuracy: 0.9838\n",
      "Epoch 15/20\n",
      "3962/3962 [==============================] - 36s 9ms/step - loss: 0.0196 - accuracy: 0.9934 - val_loss: 0.0659 - val_accuracy: 0.9839\n",
      "Epoch 16/20\n",
      "3962/3962 [==============================] - 37s 9ms/step - loss: 0.0179 - accuracy: 0.9937 - val_loss: 0.0681 - val_accuracy: 0.9838\n",
      "Epoch 17/20\n",
      "3962/3962 [==============================] - 36s 9ms/step - loss: 0.0171 - accuracy: 0.9943 - val_loss: 0.0685 - val_accuracy: 0.9840\n",
      "Epoch 18/20\n",
      "3962/3962 [==============================] - 36s 9ms/step - loss: 0.0146 - accuracy: 0.9950 - val_loss: 0.0738 - val_accuracy: 0.9825\n",
      "Epoch 19/20\n",
      "3962/3962 [==============================] - 36s 9ms/step - loss: 0.0139 - accuracy: 0.9952 - val_loss: 0.0785 - val_accuracy: 0.9836\n",
      "Epoch 20/20\n",
      "3962/3962 [==============================] - 36s 9ms/step - loss: 0.0125 - accuracy: 0.9957 - val_loss: 0.0751 - val_accuracy: 0.9829\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(train_datagen.flow(x_train, y_train, batch_size=64),\n",
    "                    validation_data=validation_datagen.flow(x_test, y_test, batch_size=32),\n",
    "                    epochs=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "afcdbfbc",
   "metadata": {},
   "outputs": [],
   "source": [
    "# model.save('ocr_model_v3.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "76d26e28",
   "metadata": {},
   "outputs": [],
   "source": [
    "import cv2\n",
    "image_h = cv2.imread('word4.jpg')\n",
    "image_h = cv2.cvtColor(image_h, cv2.COLOR_BGR2GRAY)\n",
    "_ ,binaryImg = cv2.threshold(image_h, 127, 255, cv2.THRESH_BINARY_INV)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "a22619ca",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x15c2aa91310>"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(image_h[:,75:100], cmap='gray')\n",
    "binaryImg = cv2.resize(binaryImg[:, 5:30], (28,28), interpolation = cv2.INTER_AREA)\n",
    "plt.imshow(binaryImg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "603d2b73",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 17ms/step\n"
     ]
    }
   ],
   "source": [
    "y_pred = model.predict(binaryImg.reshape(-1,28,28,1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "id": "4b6e6926",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "16\n"
     ]
    }
   ],
   "source": [
    "y_pred = np.argmax(y_pred)\n",
    "print(y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "550e98c9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['G'], dtype=object)"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "le.inverse_transform([y_pred])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "gpu2",
   "language": "python",
   "name": "gpu2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}