cnns-in-the-spotlight-digit-recognizer.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "b8e7b4dd",
   "metadata": {
    "papermill": {
     "duration": 0.009357,
     "end_time": "2024-07-18T09:48:10.595565",
     "exception": false,
     "start_time": "2024-07-18T09:48:10.586208",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **1. Import Libaries**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d2699d65",
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5",
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:10.616936Z",
     "iopub.status.busy": "2024-07-18T09:48:10.616487Z",
     "iopub.status.idle": "2024-07-18T09:48:29.614019Z",
     "shell.execute_reply": "2024-07-18T09:48:29.612658Z"
    },
    "papermill": {
     "duration": 19.010954,
     "end_time": "2024-07-18T09:48:29.616458",
     "exception": false,
     "start_time": "2024-07-18T09:48:10.605504",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-07-18 09:48:16.232423: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered\n",
      "2024-07-18 09:48:16.232613: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered\n",
      "2024-07-18 09:48:16.553324: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "from PIL import Image\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import pandas as pd\n",
    "import random\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from tensorflow.keras.utils import to_categorical\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Conv2D, MaxPool2D, Flatten, Dense, Dropout, BatchNormalization, Input\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.callbacks import LearningRateScheduler \n",
    "from PIL import Image\n",
    "import itertools\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d9609d37",
   "metadata": {
    "papermill": {
     "duration": 0.008812,
     "end_time": "2024-07-18T09:48:29.634680",
     "exception": false,
     "start_time": "2024-07-18T09:48:29.625868",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **2. Load Data**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "726bf6a9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:29.654737Z",
     "iopub.status.busy": "2024-07-18T09:48:29.654036Z",
     "iopub.status.idle": "2024-07-18T09:48:29.659691Z",
     "shell.execute_reply": "2024-07-18T09:48:29.658481Z"
    },
    "papermill": {
     "duration": 0.018291,
     "end_time": "2024-07-18T09:48:29.661954",
     "exception": false,
     "start_time": "2024-07-18T09:48:29.643663",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "train_file = \"/kaggle/input/digit-recognizer/train.csv\"\n",
    "test_file = \"/kaggle/input/digit-recognizer/test.csv\"\n",
    "output_file = \"/kaggle/input/digit-recognizer/sample_submission.csv\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "82ea0102",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:29.683198Z",
     "iopub.status.busy": "2024-07-18T09:48:29.682775Z",
     "iopub.status.idle": "2024-07-18T09:48:33.397442Z",
     "shell.execute_reply": "2024-07-18T09:48:33.396345Z"
    },
    "papermill": {
     "duration": 3.727491,
     "end_time": "2024-07-18T09:48:33.399859",
     "exception": false,
     "start_time": "2024-07-18T09:48:29.672368",
     "status": "completed"
    },
    "tags": []
   },
   "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  pixel779  \\\n",
       "0       0  ...         0         0         0         0         0         0   \n",
       "1       0  ...         0         0         0         0         0         0   \n",
       "2       0  ...         0         0         0         0         0         0   \n",
       "3       0  ...         0         0         0         0         0         0   \n",
       "4       0  ...         0         0         0         0         0         0   \n",
       "\n",
       "   pixel780  pixel781  pixel782  pixel783  \n",
       "0         0         0         0         0  \n",
       "1         0         0         0         0  \n",
       "2         0         0         0         0  \n",
       "3         0         0         0         0  \n",
       "4         0         0         0         0  \n",
       "\n",
       "[5 rows x 785 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Train Data\n",
    "import pandas as pd\n",
    "train = pd.read_csv(\"/kaggle/input/digit-recognizer/train.csv\")\n",
    "print(train.shape)\n",
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "38ca9a6a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:33.421988Z",
     "iopub.status.busy": "2024-07-18T09:48:33.421605Z",
     "iopub.status.idle": "2024-07-18T09:48:35.372016Z",
     "shell.execute_reply": "2024-07-18T09:48:35.370813Z"
    },
    "papermill": {
     "duration": 1.963722,
     "end_time": "2024-07-18T09:48:35.374603",
     "exception": false,
     "start_time": "2024-07-18T09:48:33.410881",
     "status": "completed"
    },
    "tags": []
   },
   "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  pixel779  \\\n",
       "0       0  ...         0         0         0         0         0         0   \n",
       "1       0  ...         0         0         0         0         0         0   \n",
       "2       0  ...         0         0         0         0         0         0   \n",
       "3       0  ...         0         0         0         0         0         0   \n",
       "4       0  ...         0         0         0         0         0         0   \n",
       "\n",
       "   pixel780  pixel781  pixel782  pixel783  \n",
       "0         0         0         0         0  \n",
       "1         0         0         0         0  \n",
       "2         0         0         0         0  \n",
       "3         0         0         0         0  \n",
       "4         0         0         0         0  \n",
       "\n",
       "[5 rows x 784 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Test Data\n",
    "test= pd.read_csv(\"/kaggle/input/digit-recognizer/test.csv\")\n",
    "print(test.shape)\n",
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "441fb236",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:35.397496Z",
     "iopub.status.busy": "2024-07-18T09:48:35.397103Z",
     "iopub.status.idle": "2024-07-18T09:48:35.426095Z",
     "shell.execute_reply": "2024-07-18T09:48:35.424835Z"
    },
    "papermill": {
     "duration": 0.043343,
     "end_time": "2024-07-18T09:48:35.428882",
     "exception": false,
     "start_time": "2024-07-18T09:48:35.385539",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count       785\n",
       "unique        1\n",
       "top       False\n",
       "freq        785\n",
       "dtype: object"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check the data\n",
    "train.isnull().any().describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5cddb406",
   "metadata": {
    "papermill": {
     "duration": 0.009838,
     "end_time": "2024-07-18T09:48:35.449723",
     "exception": false,
     "start_time": "2024-07-18T09:48:35.439885",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **3. Split Data**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "7bb2b3b9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:35.472234Z",
     "iopub.status.busy": "2024-07-18T09:48:35.471440Z",
     "iopub.status.idle": "2024-07-18T09:48:37.106239Z",
     "shell.execute_reply": "2024-07-18T09:48:37.105216Z"
    },
    "papermill": {
     "duration": 1.648787,
     "end_time": "2024-07-18T09:48:37.108792",
     "exception": false,
     "start_time": "2024-07-18T09:48:35.460005",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "raw_data = np.loadtxt(train_file, skiprows=1, dtype='int', delimiter=',')\n",
    "x_train, x_val, y_train, y_val = train_test_split(\n",
    "    raw_data[:,1:], raw_data[:,0], test_size=0.1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ee2522b5",
   "metadata": {
    "papermill": {
     "duration": 0.009521,
     "end_time": "2024-07-18T09:48:37.128215",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.118694",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **4. Data Visualization**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "8b8e20e3",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:37.150119Z",
     "iopub.status.busy": "2024-07-18T09:48:37.149723Z",
     "iopub.status.idle": "2024-07-18T09:48:37.482243Z",
     "shell.execute_reply": "2024-07-18T09:48:37.481050Z"
    },
    "papermill": {
     "duration": 0.346488,
     "end_time": "2024-07-18T09:48:37.484955",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.138467",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "unique_classes, counts = np.unique(y_train, return_counts=True)\n",
    "plt.figure(figsize=(8, 6))\n",
    "plt.bar(unique_classes, counts, color='lightblue', alpha=0.7)\n",
    "plt.xlabel('Class')\n",
    "plt.ylabel('Count')\n",
    "plt.title('Class Distribution in Training Data')\n",
    "plt.xticks(unique_classes)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "db565ae1",
   "metadata": {
    "papermill": {
     "duration": 0.010286,
     "end_time": "2024-07-18T09:48:37.506409",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.496123",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **5. Preprocessing**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "200dd072",
   "metadata": {
    "papermill": {
     "duration": 0.010181,
     "end_time": "2024-07-18T09:48:37.527089",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.516908",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **5.1. Reshape**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9be162dd",
   "metadata": {
    "papermill": {
     "duration": 0.010102,
     "end_time": "2024-07-18T09:48:37.547778",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.537676",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "The reshaping of x_train and x_val into 28x28 grayscale images with an additional channel (reshape(-1, 28, 28, 1)) is necessary because Convolutional Neural Networks (CNNs) expect input data to be structured in this format. CNNs are designed to process image data with specific dimensions (height x width x channels), where:\n",
    "\n",
    "* 28x28 specifies the dimensions of each image.\n",
    "* 1 indicates that each image is grayscale, meaning it has a single color channel."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "b4ef267e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:37.571116Z",
     "iopub.status.busy": "2024-07-18T09:48:37.570096Z",
     "iopub.status.idle": "2024-07-18T09:48:37.576855Z",
     "shell.execute_reply": "2024-07-18T09:48:37.575651Z"
    },
    "papermill": {
     "duration": 0.020837,
     "end_time": "2024-07-18T09:48:37.579064",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.558227",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train shape:  (37800, 28, 28, 1)\n",
      "validation shape:  (4200, 28, 28, 1)\n"
     ]
    }
   ],
   "source": [
    "x_train = x_train.reshape(-1, 28, 28, 1)\n",
    "x_val = x_val.reshape(-1, 28, 28, 1)\n",
    "print(\"train shape: \",x_train.shape)\n",
    "print(\"validation shape: \",x_val.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab847a28",
   "metadata": {
    "papermill": {
     "duration": 0.010244,
     "end_time": "2024-07-18T09:48:37.600387",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.590143",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **5.2. Normalization**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "66ede242",
   "metadata": {
    "papermill": {
     "duration": 0.010894,
     "end_time": "2024-07-18T09:48:37.621692",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.610798",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Dividing x_train and x_val by 255 converts pixel values from a range of 0-255 to a scale of 0-1. This normalization step ensures that the neural network can effectively learn from the data, leading to better training performance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "895a5ef7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:37.645423Z",
     "iopub.status.busy": "2024-07-18T09:48:37.645007Z",
     "iopub.status.idle": "2024-07-18T09:48:37.737305Z",
     "shell.execute_reply": "2024-07-18T09:48:37.736198Z"
    },
    "papermill": {
     "duration": 0.107481,
     "end_time": "2024-07-18T09:48:37.739998",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.632517",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "x_train = x_train.astype(\"float32\")/255.\n",
    "x_val = x_val.astype(\"float32\")/255."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8fbe512",
   "metadata": {
    "papermill": {
     "duration": 0.010216,
     "end_time": "2024-07-18T09:48:37.761127",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.750911",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **5.3. Encoding labels**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e58b21e8",
   "metadata": {
    "papermill": {
     "duration": 0.011231,
     "end_time": "2024-07-18T09:48:37.782838",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.771607",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Using to_categorical converts the label arrays y_train and y_val into categorical format. This transformation is crucial for training models that classify data into multiple classes. It encodes each label as a binary vector, where each position corresponds to a class, ensuring compatibility with neural networks designed for categorical classification tasks."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "cc8a7327",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:37.807045Z",
     "iopub.status.busy": "2024-07-18T09:48:37.806632Z",
     "iopub.status.idle": "2024-07-18T09:48:37.814543Z",
     "shell.execute_reply": "2024-07-18T09:48:37.813155Z"
    },
    "papermill": {
     "duration": 0.022969,
     "end_time": "2024-07-18T09:48:37.816971",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.794002",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n"
     ]
    }
   ],
   "source": [
    "y_train = to_categorical(y_train)\n",
    "y_val = to_categorical(y_val)\n",
    "#example:\n",
    "print(y_train[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c1865d1",
   "metadata": {
    "papermill": {
     "duration": 0.01126,
     "end_time": "2024-07-18T09:48:37.839061",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.827801",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **6. Convolutional Neural Network (CNN)**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "534610ec",
   "metadata": {
    "papermill": {
     "duration": 0.011262,
     "end_time": "2024-07-18T09:48:37.864396",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.853134",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "* **Input Layer:** Accepts input images of size 28x28 pixels with 1 channel (grayscale).\n",
    "\n",
    "* **Convolutional Layers:** These layers extract features from the input images using convolution operations. Each Conv2D layer is followed by Batch Normalization to stabilize and accelerate the training process by normalizing the activations. ReLU activation function is used to introduce non-linearity.\n",
    "\n",
    "* **Max Pooling Layers:** After each pair of convolutional layers, MaxPool2D layers downsample the feature maps, reducing their spatial dimensions and thereby reducing computation while preserving important features.\n",
    "\n",
    "* **Dropout Layers:** Dropout is applied after convolutional and dense layers to prevent overfitting by randomly setting a fraction of input units to zero during training.\n",
    "\n",
    "* **Flatten Layer:** Converts the 2D feature maps into a 1D feature vector, preparing the data for the dense layers.\n",
    "\n",
    "* **(Fully Connected) Layers:** These layers perform classification on the features extracted by the convolutional layers. They are densely connected, meaning each neuron in a layer is connected to every neuron in the previous layer. ReLU activation is used in the dense layers to introduce non-linearity.\n",
    "\n",
    "* **Output Layer:** The final layer has 10 neurons (corresponding to 10 classes) and uses softmax activation to output probabilities for each class, making it suitable for multi-class classification tasks."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2b01137b",
   "metadata": {
    "papermill": {
     "duration": 0.010618,
     "end_time": "2024-07-18T09:48:37.886202",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.875584",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "**CNN Architecture**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "51499168",
   "metadata": {
    "papermill": {
     "duration": 0.010894,
     "end_time": "2024-07-18T09:48:37.907950",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.897056",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "<a href=\"https://ibb.co/d4bNkDM\"><img src=\"https://i.ibb.co/jkVC3Df/cnn-1.jpg\" alt=\"cnn-1\" border=\"0\" /></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "b11202d5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:37.932506Z",
     "iopub.status.busy": "2024-07-18T09:48:37.932072Z",
     "iopub.status.idle": "2024-07-18T09:48:38.268400Z",
     "shell.execute_reply": "2024-07-18T09:48:38.267061Z"
    },
    "papermill": {
     "duration": 0.351963,
     "end_time": "2024-07-18T09:48:38.271176",
     "exception": false,
     "start_time": "2024-07-18T09:48:37.919213",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "from tensorflow.keras.layers import Input\n",
    "\n",
    "model = Sequential()\n",
    "\n",
    "model.add(Input(shape=(28, 28, 1)))\n",
    "model.add(Conv2D(filters=16, kernel_size=(3, 3), activation='relu'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Conv2D(filters=16, kernel_size=(3, 3), activation='relu'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(MaxPool2D(strides=(2, 2)))\n",
    "model.add(Dropout(0.25))\n",
    "\n",
    "model.add(Conv2D(filters=32, kernel_size=(3, 3), activation='relu'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Conv2D(filters=32, kernel_size=(3, 3), activation='relu'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(MaxPool2D(strides=(2, 2)))\n",
    "model.add(Dropout(0.25))\n",
    "\n",
    "model.add(Flatten())\n",
    "model.add(Dense(512, activation='relu'))\n",
    "model.add(Dropout(0.25))\n",
    "model.add(Dense(1024, activation='relu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(10, activation='softmax'))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2f7f13da",
   "metadata": {
    "papermill": {
     "duration": 0.010149,
     "end_time": "2024-07-18T09:48:38.292060",
     "exception": false,
     "start_time": "2024-07-18T09:48:38.281911",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **6.1 Data Augmentation**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fa2013b1",
   "metadata": {
    "papermill": {
     "duration": 0.011042,
     "end_time": "2024-07-18T09:48:38.313862",
     "exception": false,
     "start_time": "2024-07-18T09:48:38.302820",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "The ImageDataGenerator is set up to randomly zoom, shift vertically and horizontally, and rotate images slightly during training. This technique, called data augmentation, To prevent overfitting, we simulate real-world variations in handwritten digits by applying minor transformations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "e27029c5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:38.336626Z",
     "iopub.status.busy": "2024-07-18T09:48:38.336180Z",
     "iopub.status.idle": "2024-07-18T09:48:38.341482Z",
     "shell.execute_reply": "2024-07-18T09:48:38.340254Z"
    },
    "papermill": {
     "duration": 0.019388,
     "end_time": "2024-07-18T09:48:38.343776",
     "exception": false,
     "start_time": "2024-07-18T09:48:38.324388",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "datagen = ImageDataGenerator(\n",
    "    zoom_range=0.1,              #zoom images\n",
    "    height_shift_range=0.1,      #shift images vertically\n",
    "    width_shift_range=0.1,       # shift images horizontally\n",
    "    rotation_range=10            # rotate images within 10 degrees\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6134f58b",
   "metadata": {
    "papermill": {
     "duration": 0.010144,
     "end_time": "2024-07-18T09:48:38.364601",
     "exception": false,
     "start_time": "2024-07-18T09:48:38.354457",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **6.2 Training**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8b861a05",
   "metadata": {
    "papermill": {
     "duration": 0.010021,
     "end_time": "2024-07-18T09:48:38.385123",
     "exception": false,
     "start_time": "2024-07-18T09:48:38.375102",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "**Compile the model:** prepares it for training by defining how it measures performance (using categorical crossentropy for multi-class classification), optimizes weights (with Adam optimizer), and evaluates accuracy during training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "d960f65c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:38.408121Z",
     "iopub.status.busy": "2024-07-18T09:48:38.407699Z",
     "iopub.status.idle": "2024-07-18T09:48:38.425438Z",
     "shell.execute_reply": "2024-07-18T09:48:38.424094Z"
    },
    "papermill": {
     "duration": 0.032445,
     "end_time": "2024-07-18T09:48:38.427993",
     "exception": false,
     "start_time": "2024-07-18T09:48:38.395548",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "model.compile(\n",
    "    loss='categorical_crossentropy',\n",
    "    optimizer=Adam(learning_rate=1e-4),\n",
    "    metrics=['accuracy']\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cd19bc6d",
   "metadata": {
    "papermill": {
     "duration": 0.010131,
     "end_time": "2024-07-18T09:48:38.448982",
     "exception": false,
     "start_time": "2024-07-18T09:48:38.438851",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "This **LearningRateScheduler** gradually reduces the learning rate (1e-3) over epochs (x) by multiplying it with 0.9 ** x. It helps the model to stabilize and improve its performance as training progresses."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "5300f905",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:38.471564Z",
     "iopub.status.busy": "2024-07-18T09:48:38.471163Z",
     "iopub.status.idle": "2024-07-18T09:48:38.476603Z",
     "shell.execute_reply": "2024-07-18T09:48:38.475476Z"
    },
    "papermill": {
     "duration": 0.019176,
     "end_time": "2024-07-18T09:48:38.478733",
     "exception": false,
     "start_time": "2024-07-18T09:48:38.459557",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "annealer = LearningRateScheduler(lambda x: 1e-3 * 0.9 ** x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "bb7113ee",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:48:38.502133Z",
     "iopub.status.busy": "2024-07-18T09:48:38.501714Z",
     "iopub.status.idle": "2024-07-18T09:52:12.341202Z",
     "shell.execute_reply": "2024-07-18T09:52:12.339913Z"
    },
    "papermill": {
     "duration": 213.854374,
     "end_time": "2024-07-18T09:52:12.343617",
     "exception": false,
     "start_time": "2024-07-18T09:48:38.489243",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "500/500 - 15s - 31ms/step - accuracy: 0.7224 - loss: 0.8579 - val_accuracy: 0.9250 - val_loss: 0.2150 - learning_rate: 0.0010\n",
      "Epoch 2/20\n",
      "500/500 - 11s - 23ms/step - accuracy: 0.8993 - loss: 0.3386 - val_accuracy: 0.9700 - val_loss: 0.1106 - learning_rate: 9.0000e-04\n",
      "Epoch 3/20\n",
      "500/500 - 12s - 23ms/step - accuracy: 0.9209 - loss: 0.2610 - val_accuracy: 0.9600 - val_loss: 0.1091 - learning_rate: 8.1000e-04\n",
      "Epoch 4/20\n",
      "500/500 - 12s - 23ms/step - accuracy: 0.9339 - loss: 0.2133 - val_accuracy: 0.9725 - val_loss: 0.0763 - learning_rate: 7.2900e-04\n",
      "Epoch 5/20\n",
      "500/500 - 8s - 17ms/step - accuracy: 0.9471 - loss: 0.1891 - val_accuracy: 0.9675 - val_loss: 0.0955 - learning_rate: 6.5610e-04\n",
      "Epoch 6/20\n",
      "500/500 - 11s - 22ms/step - accuracy: 0.9519 - loss: 0.1616 - val_accuracy: 0.9925 - val_loss: 0.0319 - learning_rate: 5.9049e-04\n",
      "Epoch 7/20\n",
      "500/500 - 11s - 22ms/step - accuracy: 0.9524 - loss: 0.1573 - val_accuracy: 0.9725 - val_loss: 0.0770 - learning_rate: 5.3144e-04\n",
      "Epoch 8/20\n",
      "500/500 - 12s - 23ms/step - accuracy: 0.9604 - loss: 0.1348 - val_accuracy: 0.9875 - val_loss: 0.0413 - learning_rate: 4.7830e-04\n",
      "Epoch 9/20\n",
      "500/500 - 11s - 22ms/step - accuracy: 0.9604 - loss: 0.1373 - val_accuracy: 0.9900 - val_loss: 0.0454 - learning_rate: 4.3047e-04\n",
      "Epoch 10/20\n",
      "500/500 - 8s - 15ms/step - accuracy: 0.9660 - loss: 0.1136 - val_accuracy: 0.9875 - val_loss: 0.0359 - learning_rate: 3.8742e-04\n",
      "Epoch 11/20\n",
      "500/500 - 11s - 22ms/step - accuracy: 0.9643 - loss: 0.1119 - val_accuracy: 0.9900 - val_loss: 0.0460 - learning_rate: 3.4868e-04\n",
      "Epoch 12/20\n",
      "500/500 - 11s - 22ms/step - accuracy: 0.9696 - loss: 0.1086 - val_accuracy: 0.9925 - val_loss: 0.0360 - learning_rate: 3.1381e-04\n",
      "Epoch 13/20\n",
      "500/500 - 11s - 22ms/step - accuracy: 0.9720 - loss: 0.0975 - val_accuracy: 0.9900 - val_loss: 0.0307 - learning_rate: 2.8243e-04\n",
      "Epoch 14/20\n",
      "500/500 - 11s - 22ms/step - accuracy: 0.9718 - loss: 0.0919 - val_accuracy: 0.9925 - val_loss: 0.0295 - learning_rate: 2.5419e-04\n",
      "Epoch 15/20\n",
      "500/500 - 8s - 16ms/step - accuracy: 0.9742 - loss: 0.0899 - val_accuracy: 0.9900 - val_loss: 0.0333 - learning_rate: 2.2877e-04\n",
      "Epoch 16/20\n",
      "500/500 - 11s - 21ms/step - accuracy: 0.9766 - loss: 0.0762 - val_accuracy: 0.9875 - val_loss: 0.0430 - learning_rate: 2.0589e-04\n",
      "Epoch 17/20\n",
      "500/500 - 11s - 21ms/step - accuracy: 0.9770 - loss: 0.0838 - val_accuracy: 0.9925 - val_loss: 0.0329 - learning_rate: 1.8530e-04\n",
      "Epoch 18/20\n",
      "500/500 - 11s - 22ms/step - accuracy: 0.9764 - loss: 0.0763 - val_accuracy: 0.9900 - val_loss: 0.0384 - learning_rate: 1.6677e-04\n",
      "Epoch 19/20\n",
      "500/500 - 11s - 21ms/step - accuracy: 0.9765 - loss: 0.0776 - val_accuracy: 0.9950 - val_loss: 0.0174 - learning_rate: 1.5009e-04\n",
      "Epoch 20/20\n",
      "500/500 - 8s - 15ms/step - accuracy: 0.9781 - loss: 0.0766 - val_accuracy: 0.9950 - val_loss: 0.0241 - learning_rate: 1.3509e-04\n"
     ]
    }
   ],
   "source": [
    "hist = model.fit(datagen.flow(x_train, y_train, batch_size=16),\n",
    "                           steps_per_epoch=500,\n",
    "                           epochs=20,\n",
    "                           verbose=2, \n",
    "                           validation_data=(x_val[:400,:], y_val[:400,:]),\n",
    "                           callbacks=[annealer])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "f5540fc3",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:52:12.371604Z",
     "iopub.status.busy": "2024-07-18T09:52:12.371201Z",
     "iopub.status.idle": "2024-07-18T09:52:13.524651Z",
     "shell.execute_reply": "2024-07-18T09:52:13.523182Z"
    },
    "papermill": {
     "duration": 1.170362,
     "end_time": "2024-07-18T09:52:13.527211",
     "exception": false,
     "start_time": "2024-07-18T09:52:12.356849",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Final loss: 0.0230, final accuracy: 0.9931\n"
     ]
    }
   ],
   "source": [
    "final_loss, final_acc = model.evaluate(x_val, y_val, verbose=0)\n",
    "print(\"Final loss: {0:.4f}, final accuracy: {1:.4f}\".format(final_loss, final_acc))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a91cbe39",
   "metadata": {
    "papermill": {
     "duration": 0.013339,
     "end_time": "2024-07-18T09:52:13.554075",
     "exception": false,
     "start_time": "2024-07-18T09:52:13.540736",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **6.3 Evaluate The Model**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c3dbcd5b",
   "metadata": {
    "papermill": {
     "duration": 0.013003,
     "end_time": "2024-07-18T09:52:13.580402",
     "exception": false,
     "start_time": "2024-07-18T09:52:13.567399",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "* Loss Plot\n",
    "* Accuaracy Plot \n",
    "* Confusion Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "a46ea0de",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:52:13.608657Z",
     "iopub.status.busy": "2024-07-18T09:52:13.608212Z",
     "iopub.status.idle": "2024-07-18T09:52:14.117450Z",
     "shell.execute_reply": "2024-07-18T09:52:14.116332Z"
    },
    "papermill": {
     "duration": 0.526372,
     "end_time": "2024-07-18T09:52:14.119993",
     "exception": false,
     "start_time": "2024-07-18T09:52:13.593621",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(hist.history['loss'], color='b', label='Training Loss')\n",
    "plt.plot(hist.history['val_loss'], color='r', label='Validation Loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "plt.plot(hist.history['accuracy'], color='b', label='Training Accuracy')\n",
    "plt.plot(hist.history['val_accuracy'], color='r', label='Validation Accuracy')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "1ffe0d0b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:52:14.150193Z",
     "iopub.status.busy": "2024-07-18T09:52:14.149759Z",
     "iopub.status.idle": "2024-07-18T09:52:16.448510Z",
     "shell.execute_reply": "2024-07-18T09:52:16.447145Z"
    },
    "papermill": {
     "duration": 2.318097,
     "end_time": "2024-07-18T09:52:16.452354",
     "exception": false,
     "start_time": "2024-07-18T09:52:14.134257",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m132/132\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 10ms/step\n",
      "Confusion Matrix: [[394   0   0   0   0   0   0   0   1   0]\n",
      " [  0 453   1   0   1   0   0   1   0   0]\n",
      " [  0   0 439   3   0   0   0   2   0   0]\n",
      " [  0   0   0 432   0   1   0   0   0   0]\n",
      " [  0   0   0   0 421   0   0   0   0   2]\n",
      " [  0   0   0   2   0 354   1   0   2   1]\n",
      " [  0   0   0   0   0   0 419   0   0   0]\n",
      " [  0   0   0   0   1   0   0 450   0   0]\n",
      " [  0   0   0   0   1   1   0   0 417   2]\n",
      " [  1   0   0   0   3   0   0   0   2 392]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_hat = model.predict(x_val)\n",
    "y_pred = np.argmax(y_hat, axis=1)\n",
    "y_true = np.argmax(y_val, axis=1)\n",
    "cm = confusion_matrix(y_true, y_pred)\n",
    "print(\"Confusion Matrix:\", cm)\n",
    "\n",
    "plt.figure(figsize=(8, 6))\n",
    "plt.imshow(cm, interpolation='nearest', cmap=plt.cm.Blues)\n",
    "plt.title('Confusion Matrix')\n",
    "plt.colorbar()\n",
    "\n",
    "# Labeling the plot\n",
    "num_classes = len(np.unique(y_true))\n",
    "tick_marks = np.arange(num_classes)\n",
    "plt.xticks(tick_marks, range(num_classes))\n",
    "plt.yticks(tick_marks, range(num_classes))\n",
    "plt.xlabel('Predicted Label')\n",
    "plt.ylabel('True Label')\n",
    "\n",
    "thresh = cm.max() / 2.\n",
    "for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "    plt.text(j, i, format(cm[i, j], 'd'),\n",
    "             horizontalalignment=\"center\",\n",
    "             color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52df68c9",
   "metadata": {
    "papermill": {
     "duration": 0.02546,
     "end_time": "2024-07-18T09:52:16.507469",
     "exception": false,
     "start_time": "2024-07-18T09:52:16.482009",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **6.4 Test The Model**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e2d29255",
   "metadata": {
    "papermill": {
     "duration": 0.028763,
     "end_time": "2024-07-18T09:52:16.569541",
     "exception": false,
     "start_time": "2024-07-18T09:52:16.540778",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "* Testing with Test Data\n",
    "* Testing with External Images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "f5e788ca",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:52:16.627171Z",
     "iopub.status.busy": "2024-07-18T09:52:16.626077Z",
     "iopub.status.idle": "2024-07-18T09:52:27.086563Z",
     "shell.execute_reply": "2024-07-18T09:52:27.085572Z"
    },
    "papermill": {
     "duration": 10.483651,
     "end_time": "2024-07-18T09:52:27.088846",
     "exception": false,
     "start_time": "2024-07-18T09:52:16.605195",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m875/875\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 9ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABKUAAAEICAYAAACOB0fcAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuNSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/xnp5ZAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAjHklEQVR4nO3de3CV5Z3A8edguMaoQAFRMVKVqlhQUFetAi2gLaBya1ehVatu6XaLilqUWi0Kju1sVRgBt4orVikUdBjv4g2w6rhDEbGodNUVRBFBFlnAWi55948OTBHynCQn5wkhn8+Mf5hv3vc8Bn4Qf7wkuSzLsgAAAAAACTWq6wMAAAAA0PBYSgEAAACQnKUUAAAAAMlZSgEAAACQnKUUAAAAAMlZSgEAAACQnKUUAAAAAMlZSgEAAACQnKUUAAAAAMlZStUDRxxxRLj44ot3/vv8+fNDLpcL8+fPr7MzfdmXzwgNmZmF+sXMQv1iZqF+MbPEWErlMW3atJDL5Xb+06xZs9CpU6fw05/+NHzyySd1fbxqefLJJ8PYsWPr+hi7GTt27C4f4y//8/LLL9f1EalHzGzxLVu2LIwePTqccMIJoaysLLRv3z70798//OlPf6rro1EPmdl03nvvvTBs2LDQtm3b0Lx583D00UeH66+/vq6PRT1jZtMxs9QGM5vGxx9/HH70ox+Fjh07hubNm4cjjzwyXHXVVWHdunV1fbS9XkldH6C+uPnmm0PHjh3DF198EV566aVw1113hSeffDIsXbo0tGjRIulZevToEf7617+GJk2aVOu6J598MkyePHmvG+TBgweHo446are3//znPw+bNm0KJ598ch2civrOzBbP1KlTw7333huGDBkSfvKTn4QNGzaE3/72t+HUU08NTz/9dOjTp09dH5F6yMwW1+uvvx569eoVDj300HD11VeH1q1bhw8++CCsXLmyro9GPWVmi8vMUtvMbPFs2rQpnHbaaWHz5s3hJz/5SejQoUNYsmRJmDRpUpg3b15YtGhRaNTI80CVsZSqou985zvhpJNOCiGEcNlll4XWrVuH22+/PTzyyCPhggsu2OM1mzdvDqWlpbV+lkaNGoVmzZrV+n3rSpcuXUKXLl12edvKlSvDhx9+GC677LJq/2IFIZjZYrrgggvC2LFjw/7777/zbZdcckk49thjw9ixYy2lqBEzWzwVFRXhBz/4QTjmmGPCvHnzQvPmzev6SOwDzGzxmFmKwcwWz6OPPhpWrFgRHn/88dC/f/+db2/VqlW4+eabw5IlS8KJJ55Yhyfcu1nX1dC3vvWtEEII77//fgghhIsvvjjsv//+4b333gv9+vULZWVlYfjw4SGEv//GMmHChNC5c+fQrFmz0K5duzBixIiwfv36Xe6ZZVkYP358OOyww0KLFi3CN7/5zfDmm2/u9tqV/R3c//qv/wr9+vULLVu2DKWlpaFLly5h4sSJO883efLkEELY5fHNHWr7jCH8/ZHj9957r6of0l3MmDEjZFm282MIhTKztTez3bt332UhFUIIrVu3DmeeeWZ4++23814PVWFma29mn3nmmbB06dLwy1/+MjRv3jx8/vnnYfv27Xmvg+ows2aW+sXM1t7M/t///V8IIYR27drt8vb27duHEILFch6elKqhHT85W7duvfNt27ZtC2effXY444wzwm9+85udj0GOGDEiTJs2Lfzwhz8Ml19+eXj//ffDpEmTwuLFi8PLL78cGjduHEII4cYbbwzjx48P/fr1C/369QuvvfZaOOuss8KWLVvynufZZ58NAwYMCO3btw9XXHFFOPjgg8Pbb78dHn/88XDFFVeEESNGhFWrVoVnn302PPDAA7tdX4wz9u7dO4QQwvLly6v3wQ0hTJ8+PXTo0CH06NGj2tfCnpjZ4s5sCCGsXr06fOUrX6nRtfBlZrb2Zva5554LIYTQtGnTcNJJJ4VFixaFJk2ahEGDBoUpU6aEVq1a5f3vh3zMrJmlfjGztTezPXr0CI0aNQpXXHFFuO2228Jhhx0W3njjjXDLLbeEgQMHhmOOOSbvf3+DlhF13333ZSGE7LnnnsvWrl2brVy5Mps5c2bWunXrrHnz5tmHH36YZVmWXXTRRVkIIbvuuut2uf6Pf/xjFkLIpk+fvsvbn3766V3evmbNmqxJkyZZ//79s4qKip3v9/Of/zwLIWQXXXTRzrfNmzcvCyFk8+bNy7Isy7Zt25Z17NgxKy8vz9avX7/L6/zjvf7t3/4t29MPeTHOmGVZVl5enpWXl+/2evksXbo0CyFko0ePrva1YGbTz2yWZdmLL76Y5XK57IYbbqjR9TRcZrb4M3vuuedmIYSsdevW2fDhw7OHHnoou+GGG7KSkpLs9NNP3+W1IB8za2apX8xsms+Np06dmh100EFZCGHnPxdddFG2devWKl3fkPnre1XUp0+f0KZNm9ChQ4dw/vnnh/333z/MmTMnHHroobu837/+67/u8u+zZ88OBx54YOjbt2/49NNPd/6z46+/zJs3L4Tw9z8R2bJlSxg5cuQujyFeeeWVec+2ePHi8P7774crr7wyHHTQQbu0f7xXZYp1xuXLl9f4KakQgr+6R0HMbLqZXbNmTRg2bFjo2LFjGD16dLWvhxDMbE3OWNWZ3bRpUwghhJNPPjk8+OCDYciQIeHmm28O48aNC6+88kp4/vnn894DvszMmlnqFzNb3M+NDz300HDKKaeECRMmhDlz5oSrrroqTJ8+PVx33XVVur4h89f3qmjy5MmhU6dOoaSkJLRr1y587Wtf2+0r6JeUlITDDjtsl7e98847YcOGDaFt27Z7vO+aNWtCCCGsWLEihBDC0UcfvUtv06ZNaNmyZfRsOx69PP7446v+H5T4jFWVZVn4/e9/H44//vjdvvg5VIeZTTOzmzdvDgMGDAgbN24ML7300m5fawqqyswWb2Z3fC2LL38h22HDhoUxY8aEV155xTcooNrMrJmlfjGzxZvZl19+OQwYMCC8+uqrO7+Y/MCBA8MBBxwQbrrppnDJJZeE4447rsb339dZSlXRKaecsvMnWGWaNm2622BXVFSEtm3b7nz658vatGlTa2esqb3pjC+//HJYsWJFuPXWW5O9JvsmM1t8W7ZsCYMHDw5vvPFGmDt3bo0/kYAQzGwxHXLIISGE3b8A645P3r/8RWChKsxs8ZhZisHMFs9vf/vb0K5du90+vueee24YO3ZseOWVVyylIiyliuzII48Mzz33XPjGN74R/ar75eXlIYS/b3m/+tWv7nz72rVr8/7Gc+SRR4YQQli6dGn0T00qe/QxxRmravr06SGXy4Vhw4bVyv2gusxs1VRUVIQLL7wwPP/882HWrFmhZ8+eBd0PasrM5te9e/dwzz33hI8++miXt69atSqEsHf8DwUNh5nNz8yyNzGz+X3yySd7/A6ZW7duDSH8/QvIUzlfU6rIvve974Xt27eHcePG7da2bdsWPvvssxDC3/+Ob+PGjcOdd94Zsizb+T4TJkzI+xrdunULHTt2DBMmTNh5vx3+8V6lpaUhhLDb+xTrjFX9Fpo7bN26NcyePTucccYZ4fDDD6/ydVCbzGzVZnbkyJHhD3/4Q5gyZUoYPHhwla6BYjCz+Wf2vPPOC02bNg333XdfqKio2Pn2qVOnhhBC6Nu3b957QG0xs2aW+sXM5p/ZTp06hU8++STMnz9/l7fPmDEjhBDCiSeemPceDZknpYqsZ8+eYcSIEeHWW28Nr7/+ejjrrLNC48aNwzvvvBNmz54dJk6cGIYOHRratGkTrrnmmnDrrbeGAQMGhH79+oXFixeHp556Ku+3WG/UqFG46667wjnnnBNOOOGE8MMf/jC0b98+LFu2LLz55pth7ty5IYS//6lLCCFcfvnl4eyzzw777bdfOP/884t2xup+e/m5c+eGdevW+QLn1Ckzm39mJ0yYEKZMmRJOO+200KJFi/Dggw/u0gcNGrTzkwYoNjObf2YPPvjgcP3114cbb7wxfPvb3w4DBw4MS5YsCffcc0+44IILwsknn1yDjzzUjJk1s9QvZjb/zP70pz8N9913XzjnnHPCyJEjQ3l5eViwYEGYMWNG6Nu3b/inf/qnGnzkG5D03/CvftnxLTQXLlwYfb+LLrooKy0trbTffffdWffu3bPmzZtnZWVl2de//vVs9OjR2apVq3a+z/bt27Obbropa9++fda8efOsV69e2dKlS7Py8vLot9Dc4aWXXsr69u2blZWVZaWlpVmXLl2yO++8c2fftm1bNnLkyKxNmzZZLpfb7dtp1uYZs6z6317+/PPPzxo3bpytW7euytfAl5nZ4s/sjm8ZXNk/77//ft57wA5mNs3vsxUVFdmdd96ZderUKWvcuHHWoUOH7Be/+EW2ZcuWKl0PO5hZM0v9YmbTzOyyZcuyoUOHZh06dMgaN26clZeXZ9dcc022efPmKl3fkOWy7B+eWwMAAACABHxNKQAAAACSs5QCAAAAIDlLKQAAAACSs5QCAAAAIDlLKQAAAACSs5QCAAAAIDlLKQAAAACSK6nqO+ZyuWKeA9iDLMtqfK2ZhfTMLNQvZhbqFzML9UtVZtaTUgAAAAAkZykFAAAAQHKWUgAAAAAkZykFAAAAQHKWUgAAAAAkZykFAAAAQHKWUgAAAAAkZykFAAAAQHKWUgAAAAAkZykFAAAAQHKWUgAAAAAkZykFAAAAQHKWUgAAAAAkZykFAAAAQHKWUgAAAAAkV1LXBwAAANhXTJkyJdp//OMfR/vPfvazaL/tttuqfSaAvZUnpQAAAABIzlIKAAAAgOQspQAAAABIzlIKAAAAgOQspQAAAABIzlIKAAAAgOQspQAAAABILpdlWVald8zlin0W4EuqOJ57ZGYhPTML9YuZpSaGDh0a7bNmzYr2NWvWRPupp54a7cuXL4/2fZmZhfqlKjPrSSkAAAAAkrOUAgAAACA5SykAAAAAkrOUAgAAACA5SykAAAAAkrOUAgAAACA5SykAAAAAkstlWZZV6R1zuWKfBfiSKo7nHpnZ4unevXu09+jRo6D7P/HEE9H+3//93wXdn+Ixs1C/mFn2pE+fPtE+Z86caP/ss8+i/Zhjjon2zZs3R3tDZmahfqnKzHpSCgAAAIDkLKUAAAAASM5SCgAAAIDkLKUAAAAASM5SCgAAAIDkLKUAAAAASM5SCgAAAIDkSur6AAC1rXfv3tHeqlWraD/33HOjvX///tF+wAEHRHs+Y8aMifaZM2dG+0svvRTtH3zwQbS/+uqr0Q71Tb6Zb9GiRbR/+OGHtXkcYC83atSoaM/3a8aNN94Y7Zs3b672mQD2VZ6UAgAAACA5SykAAAAAkrOUAgAAACA5SykAAAAAkrOUAgAAACA5SykAAAAAkrOUAgAAACC5XJZlWZXeMZcr9lmAL6nieO5RfZ/ZQw45pNJ27rnnRq/91a9+Fe37779/jc60Q76PbSE/brUh3/k+/fTTaL/kkkui/Yknnqj2mRqKhjyze7N8v2ZMmzYt2k888cRoX7FiRXWPxF7CzDZMQ4cOjfZZs2ZF++bNm6O9rKys2meiaszs3infx3bKlCnRfvjhh0f73Llzo/3BBx+M9g0bNkT79u3bo/3ggw+O9lGjRkV7zO9///toX7JkSY3vvTeoysx6UgoAAACA5CylAAAAAEjOUgoAAACA5CylAAAAAEjOUgoAAACA5CylAAAAAEgul1Xx+2r6FprFk+9bTK5evTra8317+x/84AfRPnjw4Gjv06dPtM+ZMyfaf/zjH0f7mjVror0ha8jf9nbx4sWVtq9//esJT7K7fB/bQn7cakOh53vttdeivVevXtH++eefR/u+rCHPbF1q1qxZtP/5z3+O9g4dOkT78ccfH+3vvvtutLP3MrMNU+xzjBBC6Nq1a7SPGTMm2n/9619X+0xUjZndOzVp0iTav/jii0Qn2bOHHnoo2j/77LNov/TSS6O9kJ9b8+bNi/bevXvX+N57g6rMrCelAAAAAEjOUgoAAACA5CylAAAAAEjOUgoAAACA5CylAAAAAEjOUgoAAACA5CylAAAAAEiupK4P0BCcfvrp0T5nzpxoX7x4cbR37tw52g899NBozyfLsmgfOHBgtM+cOTPaZ82aVd0j0QB06dKl0pbv5ySF6datW7T369cv2h966KHaPA6EZs2aRfu0adOi/aijjor2MWPGRPu7774b7cDepU+fPtHetWvXaH/++eejfdKkSdU+E+zL8n1u/sknn0R7u3btavM4u/nud78b7cX+f4v33nuv0nb//fcX9bXrA09KAQAAAJCcpRQAAAAAyVlKAQAAAJCcpRQAAAAAyVlKAQAAAJCcpRQAAAAAyVlKAQAAAJBcSV0foCEYPnx4tLdp0ybazzrrrNo8DtQLf/zjHyttZ555ZvTamTNnRvvHH39cozPtkMvlov3qq68u6P7f//73o/3++++P9kaN4n/eUFFREe2rVq2K9nfeeSfaobr233//aL/qqquiffDgwdH+9NNPR/udd94Z7fu6Aw88sNKW79e7kpL4p5Ll5eXRvmjRomiHPcn3ufOIESOiPcuyaL/99tujffPmzdEODc3WrVuj/bzzzov2V199Ndrz/V60Zs2aaG/RokW0l5aWRns++c534403VtpmzJhR0GvvCzwpBQAAAEByllIAAAAAJGcpBQAAAEByllIAAAAAJGcpBQAAAEByllIAAAAAJGcpBQAAAEByJXV9gIagbdu2Rb3/okWLon3SpEnRPn369GhfvXp1tLdq1SraoSYGDBhQaTvwwAOj165duzbat2zZUqMzVfX18+ndu3e0jxgxItqzLIv2ioqKgq6/9NJLo33JkiXRDl/WvHnzaJ8/f360d+vWLdr/8z//M9ovu+yyaK/vvvrVr0b7mWeeGe133XVXpa1Zs2Y1OtMO69ati/brrrsu2u+9996CXp99U8+ePaN96NCh0Z7v15ynnnqqukcCIj744IOCrs/3ueu4ceOi/dprr4320tLSaN+2bVu05/vc/emnn472hs6TUgAAAAAkZykFAAAAQHKWUgAAAAAkZykFAAAAQHKWUgAAAAAkZykFAAAAQHKWUgAAAAAkV1LXB9gXDBkyJNoHDRpU0P0feeSRaL/kkkuiff369QW9fpZlBV0PNbFp06YatRQ2bNgQ7a1bt47222+/Pdo7d+5c7TNVx7p166L9o48+Kurr0/BMmzYt2rt16xbtL7zwQrSPGjWqukfaq+y3337RPnr06Gi/6qqror1Vq1bRvm3btkrb1q1bo9fOmjUr2ocNGxbt3bt3j/Z777032tk3lZaWRnu+mc/3uev48eOrfSZg73XLLbdEe1lZWUH3v/DCC6N95syZBd2/ofOkFAAAAADJWUoBAAAAkJylFAAAAADJWUoBAAAAkJylFAAAAADJWUoBAAAAkJylFAAAAADJldT1AfYFhx9+eLQ3ahTf/S1ZsiTaR4wYEe3r16+P9nwOPfTQaG/SpElB94d9zYABA6L92muvjfbOnTvX5nF2s3Hjxmi/8MILo/3NN9+szePQQAwcOLDSNnjw4Oi1ixYtivZrrrkm2vP9nK9rQ4YMifabb7452o899thoX7ZsWbQPGzYs2p999tloL8QLL7wQ7UcccUTRXpv6a+TIkdF+2mmnRfvMmTOj/fnnn6/2mYCaW7NmTbRPnz492ocPHx7tZWVl0b548eJoHzt2bLQ//vjj0U5hPCkFAAAAQHKWUgAAAAAkZykFAAAAQHKWUgAAAAAkZykFAAAAQHKWUgAAAAAkZykFAAAAQHIldX2AfcHSpUujfeHChdE+duzYaF+zZk11j1QtZ511VrSXlZVF+5YtW6L9tddeq/aZYG929dVXR/tpp52W6CR7NnXq1GifO3duopPQkIwePbrStt9++0Wvfeqpp6L99ddfr8mRkjnuuOOifdq0adFeWloa7RMnToz2X//619G+evXqaC+m448/PtpHjRoV7b/85S9r8zjUEyNGjIj2fJ8bjx8/vjaPA+SRy+WivWPHjtHerVu32jzOboYPHx7ty5YtK+rrE+dJKQAAAACSs5QCAAAAIDlLKQAAAACSs5QCAAAAIDlLKQAAAACSs5QCAAAAIDlLKQAAAACSK6nrA+wLnn322YJ6sZWVlUX7HXfcUdD9b7rppmh/9913C7o/1LYOHTpE+/Lly6O9UaP4Pr+ioqK6R9rFxo0bo/28886L9gULFhT0+rAnXbt2jfZTTz210rZw4cLotWPHjq3JkZJp1apVtH/ve9+L9tLS0mh/++23o/03v/lNtK9evTrai+mYY46J9n79+kX73XffXZvHoZ7o2bNntJeXl0f7U089Fe1vvfVWtc9UHd/5zneiPd/MHnvssdH+s5/9LNpvu+22aIfUOnfuHO1vvPFGopNQH3lSCgAAAIDkLKUAAAAASM5SCgAAAIDkLKUAAAAASM5SCgAAAIDkLKUAAAAASM5SCgAAAIDkSur6ABRfx44do/2AAw4o6P7PPvtsQddDal27do32LMuivaKioqDr161bF+0XXnhhtC9YsCDaoRgGDhxY42sXLVoU7flmqq59//vfj/Ybb7wx2v/nf/4n2s8+++xo/+ijj6K9mA4++OBonzdvXrS3a9cu2keNGlXtM1H/TZgwoaDrX3zxxdo5SCWmTp0a7fl+TWjSpEm05/s84frrr4/23/3ud5W2tWvXRq+FmhgwYEC0z5kzp6D75/s1oby8PNoHDRpU0OtTtzwpBQAAAEByllIAAAAAJGcpBQAAAEByllIAAAAAJGcpBQAAAEByllIAAAAAJGcpBQAAAEByJXV9AIpv8ODBBV3/9ttvR/vixYsLuj/UtrvvvjvaBw4cWNTXX7duXbRffPHF0T537txaPA3Ujt69e9f42h/96EfRXlZWFu233HJLtK9duzbaKyoqor1ly5bRfu2110b7M888E+35/vs//PDDaM8n3/kPPPDAaO/evXulbejQodFrZ8+eHe033HBDtG/atCnaqZ8OOuigaD/kkEOi/aOPPor2SZMmRfvpp58e7ffcc0+0H3vssdFeqFwuV1CH2nbYYYdF+7hx46K9UaP4sy6jRo2K9hkzZkR7p06don3QoEHRfvTRR0f7smXLop3i8qQUAAAAAMlZSgEAAACQnKUUAAAAAMlZSgEAAACQnKUUAAAAAMlZSgEAAACQnKUUAAAAAMmV1PUBKFybNm2i/dJLLy3o/n/+85+jffv27QXdH6qrRYsW0X7iiSdGe6tWrQp6/Y0bN0b7hRdeGO1z584t6PWhLvz7v/97tJ9xxhmVtkaN4n8GNnz48IL6ypUro33btm3Rvnbt2mhv3759tJeWlkb7f/zHf0T7W2+9Fe3HHXdctH/ta1+L9o4dO0Z7zMKFC6N9woQJ0b5hw4Yavzb111e+8pWC+qpVq6J9yJAh0T558uRoz/d5xMMPP1xQHzFiRLT36NEj2h966KFoz/drFlTX/fffH+1du3aN9tdeey3aJ06cWO0z/aNOnToVdP3gwYOj/bHHHivo/hTGk1IAAAAAJGcpBQAAAEByllIAAAAAJGcpBQAAAEByllIAAAAAJGcpBQAAAEByllIAAAAAJJfLsiyr0jvmcsU+CzV0wgknRPtrr71W0P27desW7a+//npB96dyVRzPPdqXZ/a73/1utM+YMaOorz9hwoRov+aaa4r6+uy99uWZbdy4cbS3adOm0nbaaadFrx04cGC0Dxs2LNobNYr/GVshPy4pbN++PdrXr18f7Rs3boz28ePHR/vcuXNr/Np//etfo31vty/PbF064ogjov0vf/lLtOf79aZQCxYsiPZFixZF+5AhQ6K9vLw82lesWBHtp5xySrSvXbs22vdlZrY48v3/XL6P+7e+9a1oz/d7ST5nnHFGtL/44ovRvnDhwmjPd/7NmzdHO5Wrysx6UgoAAACA5CylAAAAAEjOUgoAAACA5CylAAAAAEjOUgoAAACA5CylAAAAAEjOUgoAAACA5Erq+gAUrmXLlkW9/8qVK4t6f6iuXC5XUC/USSedFO39+/eP9ieeeKI2jwNJbN26NdpXrVpVaXv44Yej1+brs2bNivYmTZpEez59+vSJ9qOOOira58+fH+3Lli2L9o0bN0b7M888E+2wt1m+fHm0/+53v4v2Sy+9NNqzLIv2fJ8H9OjRI9p79uxZ0OsvWLAg2seMGRPta9eujXaornHjxkV7ly5don3y5MnRvn79+mqfKaXGjRtHe7H/34E4T0oBAAAAkJylFAAAAADJWUoBAAAAkJylFAAAAADJWUoBAAAAkJylFAAAAADJWUoBAAAAkFxJXR+Awg0ePLig61988cVo37BhQ0H3h9r2wQcfRPunn34a7a1bty7o9c8888xo7969e7T/7//+b0Gvn0++mX700UejfeTIkdFeXl5e7TPtkMvloj3Lsmh/5JFHor1nz57RPn78+GifPXt2tFM3HnvssaLe/+GHHy7q/YFd3XHHHdF+6aWXRnu+30s2b94c7Q888EC0v/rqq9H+1ltvRfuf/vSnaIfUvvGNbxR0/XPPPVdLJ6mZsrKygq6/7bbbon3Tpk0F3Z/CeFIKAAAAgOQspQAAAABIzlIKAAAAgOQspQAAAABIzlIKAAAAgOQspQAAAABILpfl+/7bO94xz7depe589tln0d68efNov+SSS6J9+vTp1T0StaSK47lHDXlm8/2c/ed//ueC7p/vY1vIj1tt2JvPtzefLYQQSkpKCrrezEL9Ymb3TlOmTIn2ESNGRPsDDzwQ7f/yL/8S7Vu3bo126o6ZrZkXXngh2nv16hXt5eXl0b5y5crqHqla97///vuj/YQTTiioL1++PNqpuarMrCelAAAAAEjOUgoAAACA5CylAAAAAEjOUgoAAACA5CylAAAAAEjOUgoAAACA5CylAAAAAEgul2VZVqV3zOWKfRYqcfzxx0f7woULo/2LL76I9pYtW1b7TKRRxfHco4Y8s61bt472gQMHRvsvfvGLaD/88MOjvZAft9qQ78e+Ls9X6NnWrFkT7X/729+qfaZ/1LFjx4KuN7NQv5hZqF/MbM288MIL0d6rV69oLy8vj/aVK1dGe9OmTaP90Ucfjfa+fftG+6RJk6L98ssvj3aKpyoz60kpAAAAAJKzlAIAAAAgOUspAAAAAJKzlAIAAAAgOUspAAAAAJKzlAIAAAAgOUspAAAAAJIrqesDkN/RRx8d7U2bNo32zz//vDaPA3u9devWRfu9995bUL/yyiureyRqyR/+8Ido//jjjxOdBACgfvjLX/4S7b169Yr2m266Kdrnzp0b7dddd120d+3aNdpnzZoV7ddff320s3fzpBQAAAAAyVlKAQAAAJCcpRQAAAAAyVlKAQAAAJCcpRQAAAAAyVlKAQAAAJCcpRQAAAAAyeWyLMuq9I65XLHPQiXuuOOOaL/iiiuifeLEidE+atSoap+JNKo4nntkZiE9Mwv1i5mF+sXM1sy3v/3taH/yyScTnWTPFixYEO35/n/3jTfeqM3jUIuqMrOelAIAAAAgOUspAAAAAJKzlAIAAAAgOUspAAAAAJKzlAIAAAAgOUspAAAAAJKzlAIAAAAguZK6PgD5tW3btqDrZ8+eXUsnAQAAoD55//33o/2xxx6L9nPOOSfaKyoqor1fv37RvmDBgmj/29/+Fu3Ub56UAgAAACA5SykAAAAAkrOUAgAAACA5SykAAAAAkrOUAgAAACA5SykAAAAAkrOUAgAAACC5XJZlWZXeMZcr9lmoxKBBg6L9V7/6VbR379492jdt2lTtM5FGFcdzj8wspGdmoX4xs1C/mFmoX6oys56UAgAAACA5SykAAAAAkrOUAgAAACA5SykAAAAAkrOUAgAAACA5SykAAAAAkrOUAgAAACC5XJZlWZXeMZcr9lmAL6nieO6RmYX0zCzUL2YW6hczC/VLVWbWk1IAAAAAJGcpBQAAAEByllIAAAAAJGcpBQAAAEByllIAAAAAJGcpBQAAAEByllIAAAAAJJfLsiyr60MAAAAA0LB4UgoAAACA5CylAAAAAEjOUgoAAACA5CylAAAAAEjOUgoAAACA5CylAAAAAEjOUgoAAACA5CylAAAAAEjOUgoAAACA5P4fAi1xUe4wx1IAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 1200x800 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_data = pd.read_csv(\"/kaggle/input/digit-recognizer/test.csv\")\n",
    "\n",
    "# Convert to numpy array and reshape\n",
    "x_test = test_data.values.reshape(-1, 28, 28, 1).astype(\"float32\") / 255.\n",
    "\n",
    "# Make predictions\n",
    "predictions = model.predict(x_test)\n",
    " \n",
    "predicted_classes = np.argmax(predictions, axis=1)\n",
    "random_indices = random.sample(range(len(x_test)), 5)\n",
    "\n",
    "plt.figure(figsize=(12, 8))\n",
    "for i, idx in enumerate(random_indices):\n",
    "    plt.subplot(2, 5, i + 1)\n",
    "    plt.imshow(x_test[idx].reshape(28, 28), cmap='gray')\n",
    "    plt.title(f'Predicted: {predicted_classes[idx]}') \n",
    "    plt.axis('off')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "9f454aa6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:52:27.142707Z",
     "iopub.status.busy": "2024-07-18T09:52:27.142336Z",
     "iopub.status.idle": "2024-07-18T09:52:27.605968Z",
     "shell.execute_reply": "2024-07-18T09:52:27.604744Z"
    },
    "papermill": {
     "duration": 0.493647,
     "end_time": "2024-07-18T09:52:27.608615",
     "exception": false,
     "start_time": "2024-07-18T09:52:27.114968",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 25ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def preprocess_image(image_path):\n",
    "    img = Image.open(image_path)\n",
    "    img = img.resize((28, 28))  \n",
    "    img = img.convert('L') \n",
    "    img_array = np.array(img).reshape(1, 28, 28, 1).astype('float32') / 255.\n",
    "    return img_array\n",
    "\n",
    "def classify_image(image_path):\n",
    "    img_array = preprocess_image(image_path)\n",
    "    if img_array is None:\n",
    "        return\n",
    "    \n",
    "    predictions = model.predict(img_array)\n",
    "    predicted_class = np.argmax(predictions)\n",
    "    \n",
    "    img = Image.open(image_path)\n",
    "    plt.imshow(img, cmap='gray')\n",
    "    plt.title(f'Predicted Class: {predicted_class}')\n",
    "    plt.axis('off')\n",
    "    plt.show()\n",
    "\n",
    "# Example usage\n",
    "image_path = '/kaggle/input/testing-model/WhatsApp Image 2024-07-18 at 10.52.00 AM.jpeg' \n",
    "classify_image(image_path)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5248170b",
   "metadata": {
    "papermill": {
     "duration": 0.026084,
     "end_time": "2024-07-18T09:52:27.662051",
     "exception": false,
     "start_time": "2024-07-18T09:52:27.635967",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **7. Submit**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "5ef58325",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-07-18T09:52:27.717391Z",
     "iopub.status.busy": "2024-07-18T09:52:27.716954Z",
     "iopub.status.idle": "2024-07-18T09:52:37.358906Z",
     "shell.execute_reply": "2024-07-18T09:52:37.357611Z"
    },
    "papermill": {
     "duration": 9.672671,
     "end_time": "2024-07-18T09:52:37.361463",
     "exception": false,
     "start_time": "2024-07-18T09:52:27.688792",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m875/875\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 9ms/step\n",
      "Submission file created.\n"
     ]
    }
   ],
   "source": [
    "test_data = pd.read_csv(\"/kaggle/input/digit-recognizer/test.csv\")\n",
    "x_test = test_data.values.reshape(-1, 28, 28, 1).astype(\"float32\") / 255.\n",
    "predictions = model.predict(x_test)\n",
    "predicted_classes = np.argmax(predictions, axis=1)\n",
    "\n",
    "# Create submission DataFrame\n",
    "submission = pd.DataFrame({\n",
    "    'ImageId': range(1, len(predicted_classes) + 1),\n",
    "    'Label': predicted_classes\n",
    "})\n",
    "\n",
    "submission.to_csv('submission.csv', index=False)\n",
    "print(\"Submission file created.\")"
   ]
  }
 ],
 "metadata": {
  "kaggle": {
   "accelerator": "none",
   "dataSources": [
    {
     "databundleVersionId": 861823,
     "sourceId": 3004,
     "sourceType": "competition"
    },
    {
     "datasetId": 5408529,
     "sourceId": 8981926,
     "sourceType": "datasetVersion"
    },
    {
     "datasetId": 5408886,
     "sourceId": 8982010,
     "sourceType": "datasetVersion"
    }
   ],
   "dockerImageVersionId": 30746,
   "isGpuEnabled": false,
   "isInternetEnabled": true,
   "language": "python",
   "sourceType": "notebook"
  },
  "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": "3.10.13"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 272.415216,
   "end_time": "2024-07-18T09:52:39.776426",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2024-07-18T09:48:07.361210",
   "version": "2.5.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}