digit-recognizer.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "207dff2c",
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5",
    "execution": {
     "iopub.execute_input": "2024-04-17T20:16:48.845014Z",
     "iopub.status.busy": "2024-04-17T20:16:48.844307Z",
     "iopub.status.idle": "2024-04-17T20:17:05.150029Z",
     "shell.execute_reply": "2024-04-17T20:17:05.148960Z"
    },
    "papermill": {
     "duration": 16.318199,
     "end_time": "2024-04-17T20:17:05.152559",
     "exception": false,
     "start_time": "2024-04-17T20:16:48.834360",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-04-17 20:16:53.903453: 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",
      "2024-04-17 20:16:53.903573: 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",
      "2024-04-17 20:16:54.068084: 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 os\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.image as mpimg\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "\n",
    "np.random.seed(2)\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import confusion_matrix\n",
    "import itertools\n",
    "\n",
    "from keras.utils import to_categorical\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPool2D\n",
    "from keras.optimizers import RMSprop\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from keras.callbacks import ReduceLROnPlateau"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "65748ee6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:05.169168Z",
     "iopub.status.busy": "2024-04-17T20:17:05.168516Z",
     "iopub.status.idle": "2024-04-17T20:17:05.175656Z",
     "shell.execute_reply": "2024-04-17T20:17:05.174877Z"
    },
    "papermill": {
     "duration": 0.017897,
     "end_time": "2024-04-17T20:17:05.178004",
     "exception": false,
     "start_time": "2024-04-17T20:17:05.160107",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/kaggle/input/digit-recognizer/sample_submission.csv\n",
      "/kaggle/input/digit-recognizer/train.csv\n",
      "/kaggle/input/digit-recognizer/test.csv\n"
     ]
    }
   ],
   "source": [
    "for dirname, _, filenames in os.walk('/kaggle/input'):\n",
    "    for filename in filenames:\n",
    "        print(os.path.join(dirname, filename))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "221e8215",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:05.194904Z",
     "iopub.status.busy": "2024-04-17T20:17:05.194281Z",
     "iopub.status.idle": "2024-04-17T20:17:10.581647Z",
     "shell.execute_reply": "2024-04-17T20:17:10.580569Z"
    },
    "papermill": {
     "duration": 5.398917,
     "end_time": "2024-04-17T20:17:10.584454",
     "exception": false,
     "start_time": "2024-04-17T20:17:05.185537",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "train_df = pd.read_csv(\"../input/digit-recognizer/train.csv\")\n",
    "test_df = pd.read_csv(\"../input/digit-recognizer/test.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "6fa6b8b4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:10.600726Z",
     "iopub.status.busy": "2024-04-17T20:17:10.600337Z",
     "iopub.status.idle": "2024-04-17T20:17:10.624150Z",
     "shell.execute_reply": "2024-04-17T20:17:10.623125Z"
    },
    "papermill": {
     "duration": 0.034868,
     "end_time": "2024-04-17T20:17:10.626706",
     "exception": false,
     "start_time": "2024-04-17T20:17:10.591838",
     "status": "completed"
    },
    "tags": []
   },
   "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>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": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "47d6e4ca",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:10.643959Z",
     "iopub.status.busy": "2024-04-17T20:17:10.643562Z",
     "iopub.status.idle": "2024-04-17T20:17:10.660386Z",
     "shell.execute_reply": "2024-04-17T20:17:10.659188Z"
    },
    "papermill": {
     "duration": 0.027971,
     "end_time": "2024-04-17T20:17:10.662441",
     "exception": false,
     "start_time": "2024-04-17T20:17:10.634470",
     "status": "completed"
    },
    "tags": []
   },
   "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>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": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "da8b6c48",
   "metadata": {
    "papermill": {
     "duration": 0.007501,
     "end_time": "2024-04-17T20:17:10.678194",
     "exception": false,
     "start_time": "2024-04-17T20:17:10.670693",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "4d5ddd3a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:10.697194Z",
     "iopub.status.busy": "2024-04-17T20:17:10.696827Z",
     "iopub.status.idle": "2024-04-17T20:17:10.715204Z",
     "shell.execute_reply": "2024-04-17T20:17:10.714006Z"
    },
    "papermill": {
     "duration": 0.030254,
     "end_time": "2024-04-17T20:17:10.717573",
     "exception": false,
     "start_time": "2024-04-17T20:17:10.687319",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df.isnull().any().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "6234f1bd",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:10.735117Z",
     "iopub.status.busy": "2024-04-17T20:17:10.734759Z",
     "iopub.status.idle": "2024-04-17T20:17:11.138557Z",
     "shell.execute_reply": "2024-04-17T20:17:11.137434Z"
    },
    "papermill": {
     "duration": 0.415425,
     "end_time": "2024-04-17T20:17:11.140920",
     "exception": false,
     "start_time": "2024-04-17T20:17:10.725495",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "label\n",
       "0    4132\n",
       "1    4684\n",
       "2    4177\n",
       "3    4351\n",
       "4    4072\n",
       "5    3795\n",
       "6    4137\n",
       "7    4401\n",
       "8    4063\n",
       "9    4188\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_train = train_df[\"label\"]\n",
    "X_train = train_df.drop(labels=[\"label\"], axis=1)\n",
    "\n",
    "del train_df\n",
    "\n",
    "y_train.value_counts().sort_index().plot(kind=\"bar\")\n",
    "\n",
    "y_train.value_counts().sort_index()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5a71bd04",
   "metadata": {
    "papermill": {
     "duration": 0.007963,
     "end_time": "2024-04-17T20:17:11.157175",
     "exception": false,
     "start_time": "2024-04-17T20:17:11.149212",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Normalization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "f3708a73",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:11.175389Z",
     "iopub.status.busy": "2024-04-17T20:17:11.174974Z",
     "iopub.status.idle": "2024-04-17T20:17:11.269930Z",
     "shell.execute_reply": "2024-04-17T20:17:11.268787Z"
    },
    "papermill": {
     "duration": 0.107056,
     "end_time": "2024-04-17T20:17:11.272415",
     "exception": false,
     "start_time": "2024-04-17T20:17:11.165359",
     "status": "completed"
    },
    "tags": []
   },
   "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>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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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     0.0     0.0     0.0     0.0     0.0   \n",
       "1     0.0     0.0     0.0     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     0.0     0.0     0.0   \n",
       "3     0.0     0.0     0.0     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     0.0     0.0     0.0   \n",
       "\n",
       "   pixel9  ...  pixel774  pixel775  pixel776  pixel777  pixel778  pixel779  \\\n",
       "0     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       0.0   \n",
       "2     0.0  ...       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       0.0   \n",
       "4     0.0  ...       0.0       0.0       0.0       0.0       0.0       0.0   \n",
       "\n",
       "   pixel780  pixel781  pixel782  pixel783  \n",
       "0       0.0       0.0       0.0       0.0  \n",
       "1       0.0       0.0       0.0       0.0  \n",
       "2       0.0       0.0       0.0       0.0  \n",
       "3       0.0       0.0       0.0       0.0  \n",
       "4       0.0       0.0       0.0       0.0  \n",
       "\n",
       "[5 rows x 784 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train = X_train / 255.0\n",
    "X_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "3de58c48",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:11.291926Z",
     "iopub.status.busy": "2024-04-17T20:17:11.290880Z",
     "iopub.status.idle": "2024-04-17T20:17:11.357991Z",
     "shell.execute_reply": "2024-04-17T20:17:11.356702Z"
    },
    "papermill": {
     "duration": 0.079466,
     "end_time": "2024-04-17T20:17:11.360449",
     "exception": false,
     "start_time": "2024-04-17T20:17:11.280983",
     "status": "completed"
    },
    "tags": []
   },
   "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>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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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     0.0     0.0     0.0     0.0     0.0   \n",
       "1     0.0     0.0     0.0     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     0.0     0.0     0.0   \n",
       "3     0.0     0.0     0.0     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     0.0     0.0     0.0   \n",
       "\n",
       "   pixel9  ...  pixel774  pixel775  pixel776  pixel777  pixel778  pixel779  \\\n",
       "0     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       0.0   \n",
       "2     0.0  ...       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       0.0   \n",
       "4     0.0  ...       0.0       0.0       0.0       0.0       0.0       0.0   \n",
       "\n",
       "   pixel780  pixel781  pixel782  pixel783  \n",
       "0       0.0       0.0       0.0       0.0  \n",
       "1       0.0       0.0       0.0       0.0  \n",
       "2       0.0       0.0       0.0       0.0  \n",
       "3       0.0       0.0       0.0       0.0  \n",
       "4       0.0       0.0       0.0       0.0  \n",
       "\n",
       "[5 rows x 784 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df = test_df / 255.0\n",
    "test_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "05c307a4",
   "metadata": {
    "papermill": {
     "duration": 0.008588,
     "end_time": "2024-04-17T20:17:11.377965",
     "exception": false,
     "start_time": "2024-04-17T20:17:11.369377",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Reshape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "fdc347a1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:11.397512Z",
     "iopub.status.busy": "2024-04-17T20:17:11.397123Z",
     "iopub.status.idle": "2024-04-17T20:17:11.402045Z",
     "shell.execute_reply": "2024-04-17T20:17:11.400944Z"
    },
    "papermill": {
     "duration": 0.017469,
     "end_time": "2024-04-17T20:17:11.404210",
     "exception": false,
     "start_time": "2024-04-17T20:17:11.386741",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "X_train = X_train.values.reshape(-1, 28, 28, 1)\n",
    "test_df = test_df.values.reshape(-1, 28, 28, 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e89dc86c",
   "metadata": {
    "papermill": {
     "duration": 0.008616,
     "end_time": "2024-04-17T20:17:11.421795",
     "exception": false,
     "start_time": "2024-04-17T20:17:11.413179",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Encoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "a058566e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:11.441510Z",
     "iopub.status.busy": "2024-04-17T20:17:11.440802Z",
     "iopub.status.idle": "2024-04-17T20:17:11.446384Z",
     "shell.execute_reply": "2024-04-17T20:17:11.445600Z"
    },
    "papermill": {
     "duration": 0.018002,
     "end_time": "2024-04-17T20:17:11.448647",
     "exception": false,
     "start_time": "2024-04-17T20:17:11.430645",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "y_train = to_categorical(y_train, num_classes=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5f1631d1",
   "metadata": {
    "papermill": {
     "duration": 0.008646,
     "end_time": "2024-04-17T20:17:11.466436",
     "exception": false,
     "start_time": "2024-04-17T20:17:11.457790",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Split training, validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "4d48f453",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:11.485996Z",
     "iopub.status.busy": "2024-04-17T20:17:11.485243Z",
     "iopub.status.idle": "2024-04-17T20:17:11.927474Z",
     "shell.execute_reply": "2024-04-17T20:17:11.926336Z"
    },
    "papermill": {
     "duration": 0.45496,
     "end_time": "2024-04-17T20:17:11.930105",
     "exception": false,
     "start_time": "2024-04-17T20:17:11.475145",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size=0.1, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "6f61e819",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:11.951236Z",
     "iopub.status.busy": "2024-04-17T20:17:11.950282Z",
     "iopub.status.idle": "2024-04-17T20:17:12.196528Z",
     "shell.execute_reply": "2024-04-17T20:17:12.195348Z"
    },
    "papermill": {
     "duration": 0.259064,
     "end_time": "2024-04-17T20:17:12.198681",
     "exception": false,
     "start_time": "2024-04-17T20:17:11.939617",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7c0da2e4f670>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_train[1][:,:,0])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "45bf8c89",
   "metadata": {
    "papermill": {
     "duration": 0.009563,
     "end_time": "2024-04-17T20:17:12.217546",
     "exception": false,
     "start_time": "2024-04-17T20:17:12.207983",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Base model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "12d8647b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:12.239110Z",
     "iopub.status.busy": "2024-04-17T20:17:12.238693Z",
     "iopub.status.idle": "2024-04-17T20:17:12.359110Z",
     "shell.execute_reply": "2024-04-17T20:17:12.358072Z"
    },
    "papermill": {
     "duration": 0.13398,
     "end_time": "2024-04-17T20:17:12.361535",
     "exception": false,
     "start_time": "2024-04-17T20:17:12.227555",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.10/site-packages/keras/src/layers/convolutional/base_conv.py:99: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
      "  super().__init__(\n"
     ]
    }
   ],
   "source": [
    "model = Sequential([\n",
    "    Conv2D(filters=32, kernel_size=(5, 5), padding='Same', activation='relu', input_shape=(28, 28, 1)),\n",
    "    Conv2D(filters=32, kernel_size=(5, 5), padding='Same', activation='relu'),\n",
    "    MaxPool2D(pool_size=(2, 2)),\n",
    "    Dropout(0.25),\n",
    "    \n",
    "    Conv2D(filters=64, kernel_size=(3, 3), padding='Same', activation='relu'),\n",
    "    Conv2D(filters=64, kernel_size=(3, 3), padding='Same', activation='relu'),\n",
    "    MaxPool2D(pool_size=(2, 2), strides=(2, 2)),\n",
    "    Dropout(0.25),\n",
    "    \n",
    "    Flatten(),\n",
    "    Dense(256, activation='relu'),\n",
    "    Dropout(0.5),\n",
    "    Dense(10, activation='softmax'),\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "48ae045a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:12.382476Z",
     "iopub.status.busy": "2024-04-17T20:17:12.382062Z",
     "iopub.status.idle": "2024-04-17T20:17:12.397055Z",
     "shell.execute_reply": "2024-04-17T20:17:12.395795Z"
    },
    "papermill": {
     "duration": 0.028273,
     "end_time": "2024-04-17T20:17:12.399463",
     "exception": false,
     "start_time": "2024-04-17T20:17:12.371190",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.10/site-packages/keras/src/optimizers/base_optimizer.py:34: UserWarning: Argument `decay` is no longer supported and will be ignored.\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "optimizer = RMSprop(learning_rate=0.001,\n",
    "                    rho=0.9,\n",
    "                    epsilon=1e-08,\n",
    "                    decay=0.0,\n",
    "                   )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "0a57be80",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:12.419956Z",
     "iopub.status.busy": "2024-04-17T20:17:12.419572Z",
     "iopub.status.idle": "2024-04-17T20:17:12.428980Z",
     "shell.execute_reply": "2024-04-17T20:17:12.427976Z"
    },
    "papermill": {
     "duration": 0.022273,
     "end_time": "2024-04-17T20:17:12.431173",
     "exception": false,
     "start_time": "2024-04-17T20:17:12.408900",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "model.compile(optimizer=optimizer,\n",
    "              loss='categorical_crossentropy',\n",
    "              metrics=['accuracy'],\n",
    "              )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "2f4878ce",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:12.451984Z",
     "iopub.status.busy": "2024-04-17T20:17:12.450990Z",
     "iopub.status.idle": "2024-04-17T20:17:12.456345Z",
     "shell.execute_reply": "2024-04-17T20:17:12.455391Z"
    },
    "papermill": {
     "duration": 0.018058,
     "end_time": "2024-04-17T20:17:12.458529",
     "exception": false,
     "start_time": "2024-04-17T20:17:12.440471",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "learning_rate_reduction = ReduceLROnPlateau(monitor='val_acc',\n",
    "                                            patience=3,\n",
    "                                            verbose=1,\n",
    "                                            factor=0.5,\n",
    "                                            min_lr=0.00001,\n",
    "                                            )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "c530555a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:12.478953Z",
     "iopub.status.busy": "2024-04-17T20:17:12.478187Z",
     "iopub.status.idle": "2024-04-17T20:17:12.482292Z",
     "shell.execute_reply": "2024-04-17T20:17:12.481548Z"
    },
    "papermill": {
     "duration": 0.016217,
     "end_time": "2024-04-17T20:17:12.484082",
     "exception": false,
     "start_time": "2024-04-17T20:17:12.467865",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "epochs = 5\n",
    "batch_size = 86"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bf6c01da",
   "metadata": {
    "papermill": {
     "duration": 0.009037,
     "end_time": "2024-04-17T20:17:12.502448",
     "exception": false,
     "start_time": "2024-04-17T20:17:12.493411",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Data augmentation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "656fb01e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:12.523161Z",
     "iopub.status.busy": "2024-04-17T20:17:12.522568Z",
     "iopub.status.idle": "2024-04-17T20:17:12.647067Z",
     "shell.execute_reply": "2024-04-17T20:17:12.645758Z"
    },
    "papermill": {
     "duration": 0.137594,
     "end_time": "2024-04-17T20:17:12.649599",
     "exception": false,
     "start_time": "2024-04-17T20:17:12.512005",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "datagen = ImageDataGenerator(\n",
    "    featurewise_center=False,\n",
    "    samplewise_center=False,\n",
    "    featurewise_std_normalization=False,\n",
    "    samplewise_std_normalization=False,\n",
    "    zca_whitening=False,\n",
    "    rotation_range=10,\n",
    "    zoom_range=0.1,\n",
    "    width_shift_range=0.1,\n",
    "    height_shift_range=0.1,\n",
    "    horizontal_flip=False,\n",
    "    vertical_flip=False,\n",
    ")\n",
    "\n",
    "datagen.fit(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "07ac565a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:17:12.670881Z",
     "iopub.status.busy": "2024-04-17T20:17:12.669652Z",
     "iopub.status.idle": "2024-04-17T20:23:15.017561Z",
     "shell.execute_reply": "2024-04-17T20:23:15.016515Z"
    },
    "papermill": {
     "duration": 362.36116,
     "end_time": "2024-04-17T20:23:15.020178",
     "exception": false,
     "start_time": "2024-04-17T20:17:12.659018",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.10/site-packages/keras/src/trainers/data_adapters/py_dataset_adapter.py:120: UserWarning: Your `PyDataset` class should call `super().__init__(**kwargs)` in its constructor. `**kwargs` can include `workers`, `use_multiprocessing`, `max_queue_size`. Do not pass these arguments to `fit()`, as they will be ignored.\n",
      "  self._warn_if_super_not_called()\n",
      "/opt/conda/lib/python3.10/site-packages/keras/src/callbacks/callback_list.py:96: UserWarning: Learning rate reduction is conditioned on metric `val_acc` which is not available. Available metrics are: accuracy,loss,val_accuracy,val_loss,learning_rate.\n",
      "  callback.on_epoch_end(epoch, logs)\n",
      "/opt/conda/lib/python3.10/contextlib.py:153: UserWarning: Your input ran out of data; interrupting training. Make sure that your dataset or generator can generate at least `steps_per_epoch * epochs` batches. You may need to use the `.repeat()` function when building your dataset.\n",
      "  self.gen.throw(typ, value, traceback)\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(datagen.flow(\n",
    "                                 X_train,\n",
    "                                 y_train,\n",
    "                                 batch_size=batch_size\n",
    "                                ),\n",
    "                    epochs = epochs,\n",
    "                    validation_data = (X_val,y_val),\n",
    "                    verbose = 0,\n",
    "                    steps_per_epoch=X_train.shape[0] // batch_size,\n",
    "                    callbacks=[learning_rate_reduction],\n",
    "                   )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "7572f93c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:23:15.041813Z",
     "iopub.status.busy": "2024-04-17T20:23:15.041412Z",
     "iopub.status.idle": "2024-04-17T20:23:15.047944Z",
     "shell.execute_reply": "2024-04-17T20:23:15.047028Z"
    },
    "papermill": {
     "duration": 0.020825,
     "end_time": "2024-04-17T20:23:15.051227",
     "exception": false,
     "start_time": "2024-04-17T20:23:15.030402",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'accuracy': [0.8656467199325562,\n",
       "  0.9651162624359131,\n",
       "  0.9612080454826355,\n",
       "  0.9767441749572754,\n",
       "  0.9719467759132385],\n",
       " 'loss': [0.4175301790237427,\n",
       "  0.046774059534072876,\n",
       "  0.12884435057640076,\n",
       "  0.08192593604326248,\n",
       "  0.09648267924785614],\n",
       " 'val_accuracy': [0.9819047451019287,\n",
       "  0.9785714149475098,\n",
       "  0.988095223903656,\n",
       "  0.9902380704879761,\n",
       "  0.9878571629524231],\n",
       " 'val_loss': [0.06446383893489838,\n",
       "  0.07361745089292526,\n",
       "  0.040085118263959885,\n",
       "  0.03541542589664459,\n",
       "  0.04041610285639763],\n",
       " 'learning_rate': [0.0010000000474974513,\n",
       "  0.0010000000474974513,\n",
       "  0.0010000000474974513,\n",
       "  0.0010000000474974513,\n",
       "  0.0010000000474974513]}"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "history.history"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "e63920e3",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:23:15.072381Z",
     "iopub.status.busy": "2024-04-17T20:23:15.071958Z",
     "iopub.status.idle": "2024-04-17T20:23:15.542755Z",
     "shell.execute_reply": "2024-04-17T20:23:15.541588Z"
    },
    "papermill": {
     "duration": 0.484292,
     "end_time": "2024-04-17T20:23:15.545242",
     "exception": false,
     "start_time": "2024-04-17T20:23:15.060950",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(2,1)\n",
    "ax[0].plot(history.history['loss'], color='b', label=\"Training loss\")\n",
    "ax[0].plot(history.history['val_loss'], color='r', label=\"validation loss\",axes =ax[0])\n",
    "legend = ax[0].legend(loc='best', shadow=True)\n",
    "\n",
    "ax[1].plot(history.history['accuracy'], color='b', label=\"Training accuracy\")\n",
    "ax[1].plot(history.history['val_accuracy'], color='r',label=\"Validation accuracy\")\n",
    "legend = ax[1].legend(loc='best', shadow=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "79e5c61c",
   "metadata": {
    "papermill": {
     "duration": 0.009824,
     "end_time": "2024-04-17T20:23:15.565320",
     "exception": false,
     "start_time": "2024-04-17T20:23:15.555496",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Submission"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "473039ca",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:23:15.587280Z",
     "iopub.status.busy": "2024-04-17T20:23:15.586612Z",
     "iopub.status.idle": "2024-04-17T20:23:37.639365Z",
     "shell.execute_reply": "2024-04-17T20:23:37.638080Z"
    },
    "papermill": {
     "duration": 22.066683,
     "end_time": "2024-04-17T20:23:37.642088",
     "exception": false,
     "start_time": "2024-04-17T20:23:15.575405",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m875/875\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m21s\u001b[0m 24ms/step\n"
     ]
    }
   ],
   "source": [
    "results = model.predict(test_df)\n",
    "\n",
    "results = np.argmax(results, axis=1)\n",
    "\n",
    "results = pd.Series(results, name=\"Label\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "3c1c8c63",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-17T20:23:37.706582Z",
     "iopub.status.busy": "2024-04-17T20:23:37.705513Z",
     "iopub.status.idle": "2024-04-17T20:23:37.737355Z",
     "shell.execute_reply": "2024-04-17T20:23:37.735828Z"
    },
    "papermill": {
     "duration": 0.066835,
     "end_time": "2024-04-17T20:23:37.739991",
     "exception": false,
     "start_time": "2024-04-17T20:23:37.673156",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "submission = pd.concat([pd.Series(range(1,28001), name=\"ImageId\"), results], axis=1)\n",
    "\n",
    "submission.to_csv(\"result.csv\", index=False)"
   ]
  }
 ],
 "metadata": {
  "kaggle": {
   "accelerator": "none",
   "dataSources": [
    {
     "databundleVersionId": 861823,
     "sourceId": 3004,
     "sourceType": "competition"
    }
   ],
   "dockerImageVersionId": 30684,
   "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": 414.301328,
   "end_time": "2024-04-17T20:23:40.329146",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2024-04-17T20:16:46.027818",
   "version": "2.5.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}