loan-approval-prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "78bba9a6",
   "metadata": {
    "papermill": {
     "duration": 0.004108,
     "end_time": "2024-12-03T09:04:24.261504",
     "exception": false,
     "start_time": "2024-12-03T09:04:24.257396",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "<div style=\"\n",
    "    background: linear-gradient(135deg, #6a11cb, #2575fc); \n",
    "    padding: 30px; \n",
    "    border-radius: 25px; \n",
    "    box-shadow: 0 12px 30px rgba(0, 0, 0, 0.4), inset 0 1px 2px rgba(255, 255, 255, 0.2); \n",
    "    text-align: center; \n",
    "    margin: 40px auto; \n",
    "    max-width: 800px; \n",
    "    border: 3px solid rgba(255, 255, 255, 0.2); \n",
    "    transition: all 0.3s ease-in-out;\">\n",
    "    <h2 style=\"\n",
    "        color: #ffffff; \n",
    "        font-size: 24px; \n",
    "        font-weight: 900; \n",
    "        font-family: 'Poppins', sans-serif; \n",
    "        margin: 0; \n",
    "        text-shadow: 3px 3px 6px rgba(0, 0, 0, 0.3), 0 0 10px rgba(255, 255, 255, 0.7); \n",
    "        letter-spacing: 1px; \n",
    "        text-transform: uppercase;\">\n",
    "        🚀 Data Preprocessing and Feature Engineering\n",
    "    </h2>\n",
    "</div>\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "10350fa3",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-12-03T09:04:24.270277Z",
     "iopub.status.busy": "2024-12-03T09:04:24.269879Z",
     "iopub.status.idle": "2024-12-03T09:04:40.895623Z",
     "shell.execute_reply": "2024-12-03T09:04:40.894666Z"
    },
    "papermill": {
     "duration": 16.633094,
     "end_time": "2024-12-03T09:04:40.898211",
     "exception": false,
     "start_time": "2024-12-03T09:04:24.265117",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import roc_auc_score, accuracy_score\n",
    "from sklearn.preprocessing import LabelEncoder, StandardScaler\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "import xgboost as xgb\n",
    "import catboost as cb\n",
    "from sklearn.ensemble import RandomForestClassifier, ExtraTreesClassifier\n",
    "from sklearn.impute import SimpleImputer\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras import layers, models\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "df = pd.read_csv('/kaggle/input/playground-series-s4e10/train.csv')\n",
    "\n",
    "numerical_cols = df.select_dtypes(include=[np.number]).columns.tolist()\n",
    "categorical_cols = df.select_dtypes(include=[object]).columns.tolist()\n",
    "\n",
    "numerical_cols.remove('loan_status')\n",
    "\n",
    "scaler = StandardScaler()\n",
    "df[numerical_cols] = scaler.fit_transform(df[numerical_cols])\n",
    "\n",
    "label_encoder = LabelEncoder()\n",
    "for col in categorical_cols:\n",
    "    df[col] = label_encoder.fit_transform(df[col])\n",
    "\n",
    "y = df['loan_status']\n",
    "if not np.array_equal(np.unique(y), [0, 1]):\n",
    "    label_encoder = LabelEncoder()\n",
    "    y = label_encoder.fit_transform(y)\n",
    "\n",
    "X = df.drop('loan_status', axis=1)\n",
    "X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "18f434b0",
   "metadata": {
    "papermill": {
     "duration": 0.003227,
     "end_time": "2024-12-03T09:04:40.905126",
     "exception": false,
     "start_time": "2024-12-03T09:04:40.901899",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "<div style=\"\n",
    "    background: linear-gradient(135deg, #6a11cb, #2575fc); \n",
    "    padding: 30px; \n",
    "    border-radius: 25px; \n",
    "    box-shadow: 0 12px 30px rgba(0, 0, 0, 0.4), inset 0 1px 2px rgba(255, 255, 255, 0.2); \n",
    "    text-align: center; \n",
    "    margin: 40px auto; \n",
    "    max-width: 800px; \n",
    "    border: 3px solid rgba(255, 255, 255, 0.2); \n",
    "    transition: all 0.3s ease-in-out;\">\n",
    "    <h2 style=\"\n",
    "        color: #ffffff; \n",
    "        font-size: 24px; \n",
    "        font-weight: 900; \n",
    "        font-family: 'Poppins', sans-serif; \n",
    "        margin: 0; \n",
    "        text-shadow: 3px 3px 6px rgba(0, 0, 0, 0.3), 0 0 10px rgba(255, 255, 255, 0.7); \n",
    "        letter-spacing: 1px; \n",
    "        text-transform: uppercase;\">\n",
    "        🔹 XGBoost Model Training and Evaluation\n",
    "    </h2>\n",
    "</div>\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "946e254c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-12-03T09:04:40.913362Z",
     "iopub.status.busy": "2024-12-03T09:04:40.912978Z",
     "iopub.status.idle": "2024-12-03T09:04:41.338691Z",
     "shell.execute_reply": "2024-12-03T09:04:41.337268Z"
    },
    "papermill": {
     "duration": 0.433557,
     "end_time": "2024-12-03T09:04:41.341939",
     "exception": false,
     "start_time": "2024-12-03T09:04:40.908382",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "XGBoost Accuracy: 0.9525961292522807\n",
      "XGBoost AUC-ROC: 0.9544392530417765\n"
     ]
    }
   ],
   "source": [
    "xgb_model = xgb.XGBClassifier(eval_metric=\"logloss\", random_state=42)\n",
    "xgb_model.fit(X_train, y_train)\n",
    "\n",
    "y_pred_xgb = xgb_model.predict(X_val)\n",
    "y_pred_proba_xgb = xgb_model.predict_proba(X_val)[:, 1]\n",
    "\n",
    "print(\"XGBoost Accuracy:\", accuracy_score(y_val, y_pred_xgb))\n",
    "print(\"XGBoost AUC-ROC:\", roc_auc_score(y_val, y_pred_proba_xgb))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "687c000d",
   "metadata": {
    "papermill": {
     "duration": 0.005371,
     "end_time": "2024-12-03T09:04:41.353309",
     "exception": false,
     "start_time": "2024-12-03T09:04:41.347938",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "<div style=\"\n",
    "    background: linear-gradient(135deg, #6a11cb, #2575fc); \n",
    "    padding: 30px; \n",
    "    border-radius: 25px; \n",
    "    box-shadow: 0 12px 30px rgba(0, 0, 0, 0.4), inset 0 1px 2px rgba(255, 255, 255, 0.2); \n",
    "    text-align: center; \n",
    "    margin: 40px auto; \n",
    "    max-width: 800px; \n",
    "    border: 3px solid rgba(255, 255, 255, 0.2); \n",
    "    transition: all 0.3s ease-in-out;\">\n",
    "    <h2 style=\"\n",
    "        color: #ffffff; \n",
    "        font-size: 24px; \n",
    "        font-weight: 900; \n",
    "        font-family: 'Poppins', sans-serif; \n",
    "        margin: 0; \n",
    "        text-shadow: 3px 3px 6px rgba(0, 0, 0, 0.3), 0 0 10px rgba(255, 255, 255, 0.7); \n",
    "        letter-spacing: 1px; \n",
    "        text-transform: uppercase;\">\n",
    "        🔹 CatBoost Model Training and Evaluation\n",
    "    </h2>\n",
    "</div>\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "d7556e90",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-12-03T09:04:41.365071Z",
     "iopub.status.busy": "2024-12-03T09:04:41.364317Z",
     "iopub.status.idle": "2024-12-03T09:04:56.701933Z",
     "shell.execute_reply": "2024-12-03T09:04:56.700829Z"
    },
    "papermill": {
     "duration": 15.346939,
     "end_time": "2024-12-03T09:04:56.704247",
     "exception": false,
     "start_time": "2024-12-03T09:04:41.357308",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CatBoost Accuracy: 0.9533634580953193\n",
      "CatBoost AUC-ROC: 0.9526089525392182\n"
     ]
    }
   ],
   "source": [
    "cat_model = cb.CatBoostClassifier(iterations=500, learning_rate=0.05, depth=10, random_state=42, verbose=0)\n",
    "cat_model.fit(X_train, y_train)\n",
    "\n",
    "y_pred_cat = cat_model.predict(X_val)\n",
    "y_pred_proba_cat = cat_model.predict_proba(X_val)[:, 1]\n",
    "\n",
    "print(\"CatBoost Accuracy:\", accuracy_score(y_val, y_pred_cat))\n",
    "print(\"CatBoost AUC-ROC:\", roc_auc_score(y_val, y_pred_proba_cat))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0ba2bcb5",
   "metadata": {
    "papermill": {
     "duration": 0.003467,
     "end_time": "2024-12-03T09:04:56.711610",
     "exception": false,
     "start_time": "2024-12-03T09:04:56.708143",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "<div style=\"\n",
    "    background: linear-gradient(135deg, #6a11cb, #2575fc); \n",
    "    padding: 30px; \n",
    "    border-radius: 25px; \n",
    "    box-shadow: 0 12px 30px rgba(0, 0, 0, 0.4), inset 0 1px 2px rgba(255, 255, 255, 0.2); \n",
    "    text-align: center; \n",
    "    margin: 40px auto; \n",
    "    max-width: 800px; \n",
    "    border: 3px solid rgba(255, 255, 255, 0.2); \n",
    "    transition: all 0.3s ease-in-out;\">\n",
    "    <h2 style=\"\n",
    "        color: #ffffff; \n",
    "        font-size: 24px; \n",
    "        font-weight: 900; \n",
    "        font-family: 'Poppins', sans-serif; \n",
    "        margin: 0; \n",
    "        text-shadow: 3px 3px 6px rgba(0, 0, 0, 0.3), 0 0 10px rgba(255, 255, 255, 0.7); \n",
    "        letter-spacing: 1px; \n",
    "        text-transform: uppercase;\">\n",
    "        🔹 Random Forest Model Training and Evaluation\n",
    "    </h2>\n",
    "</div>\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "95c6ea7d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-12-03T09:04:56.722070Z",
     "iopub.status.busy": "2024-12-03T09:04:56.721301Z",
     "iopub.status.idle": "2024-12-03T09:05:04.397111Z",
     "shell.execute_reply": "2024-12-03T09:05:04.395970Z"
    },
    "papermill": {
     "duration": 7.684016,
     "end_time": "2024-12-03T09:05:04.399280",
     "exception": false,
     "start_time": "2024-12-03T09:04:56.715264",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest Accuracy: 0.949356296359451\n",
      "Random Forest AUC-ROC: 0.936624569654481\n"
     ]
    }
   ],
   "source": [
    "rf_model = RandomForestClassifier(n_estimators=100, random_state=42)\n",
    "rf_model.fit(X_train, y_train)\n",
    "\n",
    "y_pred_rf = rf_model.predict(X_val)\n",
    "y_pred_proba_rf = rf_model.predict_proba(X_val)[:, 1]\n",
    "\n",
    "print(\"Random Forest Accuracy:\", accuracy_score(y_val, y_pred_rf))\n",
    "print(\"Random Forest AUC-ROC:\", roc_auc_score(y_val, y_pred_proba_rf))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "33b22d9e",
   "metadata": {
    "papermill": {
     "duration": 0.003199,
     "end_time": "2024-12-03T09:05:04.406107",
     "exception": false,
     "start_time": "2024-12-03T09:05:04.402908",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "<div style=\"\n",
    "    background: linear-gradient(135deg, #6a11cb, #2575fc); \n",
    "    padding: 30px; \n",
    "    border-radius: 25px; \n",
    "    box-shadow: 0 12px 30px rgba(0, 0, 0, 0.4), inset 0 1px 2px rgba(255, 255, 255, 0.2); \n",
    "    text-align: center; \n",
    "    margin: 40px auto; \n",
    "    max-width: 800px; \n",
    "    border: 3px solid rgba(255, 255, 255, 0.2); \n",
    "    transition: all 0.3s ease-in-out;\">\n",
    "    <h2 style=\"\n",
    "        color: #ffffff; \n",
    "        font-size: 24px; \n",
    "        font-weight: 900; \n",
    "        font-family: 'Poppins', sans-serif; \n",
    "        margin: 0; \n",
    "        text-shadow: 3px 3px 6px rgba(0, 0, 0, 0.3), 0 0 10px rgba(255, 255, 255, 0.7); \n",
    "        letter-spacing: 1px; \n",
    "        text-transform: uppercase;\">\n",
    "        🔹 Extra Trees Classifier Model Training and Evaluation\n",
    "    </h2>\n",
    "</div>\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "803d47e9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-12-03T09:05:04.414958Z",
     "iopub.status.busy": "2024-12-03T09:05:04.414184Z",
     "iopub.status.idle": "2024-12-03T09:05:08.780851Z",
     "shell.execute_reply": "2024-12-03T09:05:08.779653Z"
    },
    "papermill": {
     "duration": 4.37349,
     "end_time": "2024-12-03T09:05:08.783020",
     "exception": false,
     "start_time": "2024-12-03T09:05:04.409530",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extra Trees Accuracy: 0.9487594850370875\n",
      "Extra Trees AUC-ROC: 0.926066817952993\n"
     ]
    }
   ],
   "source": [
    "et_model = ExtraTreesClassifier(n_estimators=100, random_state=42)\n",
    "et_model.fit(X_train, y_train)\n",
    "\n",
    "y_pred_et = et_model.predict(X_val)\n",
    "y_pred_proba_et = et_model.predict_proba(X_val)[:, 1]\n",
    "\n",
    "print(\"Extra Trees Accuracy:\", accuracy_score(y_val, y_pred_et))\n",
    "print(\"Extra Trees AUC-ROC:\", roc_auc_score(y_val, y_pred_proba_et))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7fa67d15",
   "metadata": {
    "papermill": {
     "duration": 0.003991,
     "end_time": "2024-12-03T09:05:08.790825",
     "exception": false,
     "start_time": "2024-12-03T09:05:08.786834",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "<div style=\"\n",
    "    background: linear-gradient(135deg, #6a11cb, #2575fc); \n",
    "    padding: 30px; \n",
    "    border-radius: 25px; \n",
    "    box-shadow: 0 12px 30px rgba(0, 0, 0, 0.4), inset 0 1px 2px rgba(255, 255, 255, 0.2); \n",
    "    text-align: center; \n",
    "    margin: 40px auto; \n",
    "    max-width: 800px; \n",
    "    border: 3px solid rgba(255, 255, 255, 0.2); \n",
    "    transition: all 0.3s ease-in-out;\">\n",
    "    <h2 style=\"\n",
    "        color: #ffffff; \n",
    "        font-size: 24px; \n",
    "        font-weight: 900; \n",
    "        font-family: 'Poppins', sans-serif; \n",
    "        margin: 0; \n",
    "        text-shadow: 3px 3px 6px rgba(0, 0, 0, 0.3), 0 0 10px rgba(255, 255, 255, 0.7); \n",
    "        letter-spacing: 1px; \n",
    "        text-transform: uppercase;\">\n",
    "        🔹 Neural Network Model Training and Evaluation\n",
    "    </h2>\n",
    "</div>\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "837a1410",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-12-03T09:05:08.805840Z",
     "iopub.status.busy": "2024-12-03T09:05:08.804864Z",
     "iopub.status.idle": "2024-12-03T09:09:52.030117Z",
     "shell.execute_reply": "2024-12-03T09:09:52.028924Z"
    },
    "papermill": {
     "duration": 283.235022,
     "end_time": "2024-12-03T09:09:52.032232",
     "exception": false,
     "start_time": "2024-12-03T09:05:08.797210",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 2ms/step - accuracy: 0.8593 - loss: 0.3480 - val_accuracy: 0.9020 - val_loss: 0.2516\n",
      "Epoch 2/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.8991 - loss: 0.2649 - val_accuracy: 0.9135 - val_loss: 0.2342\n",
      "Epoch 3/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9057 - loss: 0.2513 - val_accuracy: 0.9178 - val_loss: 0.2264\n",
      "Epoch 4/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9125 - loss: 0.2385 - val_accuracy: 0.9178 - val_loss: 0.2248\n",
      "Epoch 5/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9151 - loss: 0.2355 - val_accuracy: 0.9250 - val_loss: 0.2176\n",
      "Epoch 6/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9167 - loss: 0.2294 - val_accuracy: 0.9304 - val_loss: 0.2110\n",
      "Epoch 7/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9193 - loss: 0.2298 - val_accuracy: 0.9310 - val_loss: 0.2124\n",
      "Epoch 8/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9203 - loss: 0.2246 - val_accuracy: 0.9225 - val_loss: 0.2177\n",
      "Epoch 9/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9228 - loss: 0.2224 - val_accuracy: 0.9315 - val_loss: 0.2117\n",
      "Epoch 10/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9197 - loss: 0.2284 - val_accuracy: 0.9344 - val_loss: 0.2087\n",
      "Epoch 11/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9242 - loss: 0.2184 - val_accuracy: 0.9339 - val_loss: 0.2071\n",
      "Epoch 12/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9237 - loss: 0.2196 - val_accuracy: 0.9315 - val_loss: 0.2101\n",
      "Epoch 13/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9243 - loss: 0.2176 - val_accuracy: 0.9317 - val_loss: 0.2086\n",
      "Epoch 14/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9265 - loss: 0.2143 - val_accuracy: 0.9315 - val_loss: 0.2096\n",
      "Epoch 15/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9276 - loss: 0.2127 - val_accuracy: 0.9321 - val_loss: 0.2102\n",
      "Epoch 16/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9274 - loss: 0.2120 - val_accuracy: 0.9376 - val_loss: 0.2035\n",
      "Epoch 17/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9272 - loss: 0.2136 - val_accuracy: 0.9329 - val_loss: 0.2056\n",
      "Epoch 18/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9254 - loss: 0.2165 - val_accuracy: 0.9322 - val_loss: 0.2067\n",
      "Epoch 19/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9286 - loss: 0.2124 - val_accuracy: 0.9363 - val_loss: 0.1994\n",
      "Epoch 20/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9292 - loss: 0.2134 - val_accuracy: 0.9350 - val_loss: 0.2025\n",
      "Epoch 21/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9285 - loss: 0.2096 - val_accuracy: 0.9350 - val_loss: 0.2010\n",
      "Epoch 22/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9290 - loss: 0.2068 - val_accuracy: 0.9351 - val_loss: 0.1995\n",
      "Epoch 23/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9300 - loss: 0.2082 - val_accuracy: 0.9365 - val_loss: 0.1995\n",
      "Epoch 24/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9307 - loss: 0.2081 - val_accuracy: 0.9370 - val_loss: 0.1985\n",
      "Epoch 25/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9310 - loss: 0.2091 - val_accuracy: 0.9354 - val_loss: 0.2008\n",
      "Epoch 26/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9301 - loss: 0.2080 - val_accuracy: 0.9376 - val_loss: 0.1943\n",
      "Epoch 27/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9314 - loss: 0.2054 - val_accuracy: 0.9362 - val_loss: 0.2009\n",
      "Epoch 28/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9331 - loss: 0.2020 - val_accuracy: 0.9368 - val_loss: 0.1965\n",
      "Epoch 29/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9332 - loss: 0.2029 - val_accuracy: 0.9392 - val_loss: 0.1959\n",
      "Epoch 30/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9328 - loss: 0.2015 - val_accuracy: 0.9396 - val_loss: 0.1967\n",
      "Epoch 31/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9333 - loss: 0.2031 - val_accuracy: 0.9364 - val_loss: 0.1948\n",
      "Epoch 32/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9320 - loss: 0.2044 - val_accuracy: 0.9390 - val_loss: 0.1923\n",
      "Epoch 33/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9317 - loss: 0.2060 - val_accuracy: 0.9334 - val_loss: 0.2014\n",
      "Epoch 34/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9327 - loss: 0.2031 - val_accuracy: 0.9383 - val_loss: 0.1930\n",
      "Epoch 35/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9323 - loss: 0.2048 - val_accuracy: 0.9381 - val_loss: 0.1948\n",
      "Epoch 36/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9337 - loss: 0.2006 - val_accuracy: 0.9372 - val_loss: 0.1919\n",
      "Epoch 37/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9316 - loss: 0.2038 - val_accuracy: 0.9384 - val_loss: 0.1932\n",
      "Epoch 38/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9331 - loss: 0.2018 - val_accuracy: 0.9390 - val_loss: 0.1941\n",
      "Epoch 39/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9321 - loss: 0.2036 - val_accuracy: 0.9373 - val_loss: 0.1940\n",
      "Epoch 40/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9341 - loss: 0.2039 - val_accuracy: 0.9393 - val_loss: 0.1921\n",
      "Epoch 41/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9343 - loss: 0.1981 - val_accuracy: 0.9381 - val_loss: 0.1924\n",
      "Epoch 42/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9332 - loss: 0.2037 - val_accuracy: 0.9396 - val_loss: 0.1930\n",
      "Epoch 43/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9366 - loss: 0.1978 - val_accuracy: 0.9377 - val_loss: 0.1929\n",
      "Epoch 44/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9365 - loss: 0.1970 - val_accuracy: 0.9364 - val_loss: 0.1946\n",
      "Epoch 45/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9351 - loss: 0.1992 - val_accuracy: 0.9392 - val_loss: 0.1900\n",
      "Epoch 46/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9357 - loss: 0.2002 - val_accuracy: 0.9404 - val_loss: 0.1905\n",
      "Epoch 47/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9360 - loss: 0.1982 - val_accuracy: 0.9411 - val_loss: 0.1893\n",
      "Epoch 48/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9345 - loss: 0.1979 - val_accuracy: 0.9378 - val_loss: 0.1936\n",
      "Epoch 49/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9375 - loss: 0.1962 - val_accuracy: 0.9397 - val_loss: 0.1918\n",
      "Epoch 50/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9337 - loss: 0.2000 - val_accuracy: 0.9378 - val_loss: 0.1929\n",
      "Epoch 51/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9355 - loss: 0.1989 - val_accuracy: 0.9402 - val_loss: 0.1913\n",
      "Epoch 52/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9364 - loss: 0.1986 - val_accuracy: 0.9393 - val_loss: 0.1906\n",
      "Epoch 53/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9346 - loss: 0.2008 - val_accuracy: 0.9396 - val_loss: 0.1885\n",
      "Epoch 54/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9349 - loss: 0.2000 - val_accuracy: 0.9400 - val_loss: 0.1890\n",
      "Epoch 55/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9383 - loss: 0.1928 - val_accuracy: 0.9399 - val_loss: 0.1880\n",
      "Epoch 56/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9374 - loss: 0.1926 - val_accuracy: 0.9364 - val_loss: 0.1969\n",
      "Epoch 57/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9363 - loss: 0.1974 - val_accuracy: 0.9409 - val_loss: 0.1895\n",
      "Epoch 58/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9363 - loss: 0.1978 - val_accuracy: 0.9404 - val_loss: 0.1887\n",
      "Epoch 59/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9360 - loss: 0.1986 - val_accuracy: 0.9407 - val_loss: 0.1876\n",
      "Epoch 60/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9372 - loss: 0.1935 - val_accuracy: 0.9394 - val_loss: 0.1911\n",
      "Epoch 61/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9360 - loss: 0.1965 - val_accuracy: 0.9408 - val_loss: 0.1896\n",
      "Epoch 62/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9395 - loss: 0.1957 - val_accuracy: 0.9384 - val_loss: 0.1913\n",
      "Epoch 63/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9360 - loss: 0.1996 - val_accuracy: 0.9407 - val_loss: 0.1881\n",
      "Epoch 64/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9363 - loss: 0.1972 - val_accuracy: 0.9410 - val_loss: 0.1875\n",
      "Epoch 65/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9371 - loss: 0.1963 - val_accuracy: 0.9419 - val_loss: 0.1865\n",
      "Epoch 66/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9376 - loss: 0.1989 - val_accuracy: 0.9418 - val_loss: 0.1880\n",
      "Epoch 67/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9381 - loss: 0.1950 - val_accuracy: 0.9426 - val_loss: 0.1855\n",
      "Epoch 68/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9364 - loss: 0.1968 - val_accuracy: 0.9413 - val_loss: 0.1860\n",
      "Epoch 69/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9381 - loss: 0.1964 - val_accuracy: 0.9435 - val_loss: 0.1848\n",
      "Epoch 70/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9382 - loss: 0.1945 - val_accuracy: 0.9419 - val_loss: 0.1862\n",
      "Epoch 71/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9395 - loss: 0.1924 - val_accuracy: 0.9414 - val_loss: 0.1844\n",
      "Epoch 72/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9383 - loss: 0.1952 - val_accuracy: 0.9441 - val_loss: 0.1858\n",
      "Epoch 73/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9402 - loss: 0.1915 - val_accuracy: 0.9438 - val_loss: 0.1854\n",
      "Epoch 74/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9395 - loss: 0.1955 - val_accuracy: 0.9437 - val_loss: 0.1838\n",
      "Epoch 75/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9366 - loss: 0.2026 - val_accuracy: 0.9442 - val_loss: 0.1833\n",
      "Epoch 76/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9386 - loss: 0.1953 - val_accuracy: 0.9427 - val_loss: 0.1838\n",
      "Epoch 77/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9415 - loss: 0.1909 - val_accuracy: 0.9434 - val_loss: 0.1849\n",
      "Epoch 78/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9398 - loss: 0.1920 - val_accuracy: 0.9426 - val_loss: 0.1845\n",
      "Epoch 79/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9410 - loss: 0.1898 - val_accuracy: 0.9435 - val_loss: 0.1819\n",
      "Epoch 80/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9393 - loss: 0.1943 - val_accuracy: 0.9443 - val_loss: 0.1824\n",
      "Epoch 81/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9408 - loss: 0.1907 - val_accuracy: 0.9446 - val_loss: 0.1826\n",
      "Epoch 82/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9402 - loss: 0.1924 - val_accuracy: 0.9441 - val_loss: 0.1816\n",
      "Epoch 83/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9416 - loss: 0.1902 - val_accuracy: 0.9465 - val_loss: 0.1814\n",
      "Epoch 84/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9392 - loss: 0.1948 - val_accuracy: 0.9435 - val_loss: 0.1830\n",
      "Epoch 85/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9399 - loss: 0.1924 - val_accuracy: 0.9439 - val_loss: 0.1811\n",
      "Epoch 86/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9396 - loss: 0.1934 - val_accuracy: 0.9438 - val_loss: 0.1832\n",
      "Epoch 87/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9381 - loss: 0.1950 - val_accuracy: 0.9442 - val_loss: 0.1819\n",
      "Epoch 88/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9401 - loss: 0.1934 - val_accuracy: 0.9448 - val_loss: 0.1821\n",
      "Epoch 89/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9396 - loss: 0.1945 - val_accuracy: 0.9436 - val_loss: 0.1809\n",
      "Epoch 90/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9419 - loss: 0.1887 - val_accuracy: 0.9443 - val_loss: 0.1829\n",
      "Epoch 91/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9426 - loss: 0.1845 - val_accuracy: 0.9448 - val_loss: 0.1814\n",
      "Epoch 92/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9401 - loss: 0.1936 - val_accuracy: 0.9448 - val_loss: 0.1814\n",
      "Epoch 93/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9414 - loss: 0.1901 - val_accuracy: 0.9456 - val_loss: 0.1813\n",
      "Epoch 94/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9403 - loss: 0.1935 - val_accuracy: 0.9468 - val_loss: 0.1808\n",
      "Epoch 95/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9416 - loss: 0.1881 - val_accuracy: 0.9471 - val_loss: 0.1801\n",
      "Epoch 96/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9397 - loss: 0.1933 - val_accuracy: 0.9469 - val_loss: 0.1796\n",
      "Epoch 97/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9403 - loss: 0.1940 - val_accuracy: 0.9465 - val_loss: 0.1784\n",
      "Epoch 98/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9418 - loss: 0.1889 - val_accuracy: 0.9453 - val_loss: 0.1820\n",
      "Epoch 99/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9407 - loss: 0.1918 - val_accuracy: 0.9471 - val_loss: 0.1797\n",
      "Epoch 100/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9410 - loss: 0.1908 - val_accuracy: 0.9438 - val_loss: 0.1812\n",
      "Epoch 101/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9408 - loss: 0.1945 - val_accuracy: 0.9472 - val_loss: 0.1791\n",
      "Epoch 102/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9378 - loss: 0.1963 - val_accuracy: 0.9449 - val_loss: 0.1815\n",
      "Epoch 103/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9406 - loss: 0.1877 - val_accuracy: 0.9480 - val_loss: 0.1790\n",
      "Epoch 104/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9391 - loss: 0.1947 - val_accuracy: 0.9464 - val_loss: 0.1808\n",
      "Epoch 105/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9408 - loss: 0.1889 - val_accuracy: 0.9463 - val_loss: 0.1816\n",
      "Epoch 106/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9400 - loss: 0.1943 - val_accuracy: 0.9454 - val_loss: 0.1801\n",
      "Epoch 107/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9438 - loss: 0.1846 - val_accuracy: 0.9433 - val_loss: 0.1812\n",
      "Epoch 108/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9416 - loss: 0.1901 - val_accuracy: 0.9477 - val_loss: 0.1784\n",
      "Epoch 109/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9416 - loss: 0.1917 - val_accuracy: 0.9471 - val_loss: 0.1793\n",
      "Epoch 110/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 2ms/step - accuracy: 0.9408 - loss: 0.1938 - val_accuracy: 0.9453 - val_loss: 0.1807\n",
      "Epoch 111/111\n",
      "\u001b[1m1467/1467\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 2ms/step - accuracy: 0.9430 - loss: 0.1878 - val_accuracy: 0.9468 - val_loss: 0.1796\n",
      "\u001b[1m367/367\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step  \n",
      "Neural Network Accuracy: 0.9467985335493222\n",
      "Neural Network AUC-ROC: 0.9314496764869145\n"
     ]
    }
   ],
   "source": [
    "nn_model = models.Sequential([\n",
    "    layers.Dense(64, activation='relu', input_shape=(X_train.shape[1],)),\n",
    "    layers.Dropout(0.5),\n",
    "    layers.Dense(32, activation='relu'),\n",
    "    layers.Dense(1, activation='sigmoid') \n",
    "])\n",
    "nn_model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
    "history = nn_model.fit(X_train, y_train, epochs=111, batch_size=32, validation_data=(X_val, y_val), verbose=1)\n",
    "y_pred_nn = nn_model.predict(X_val).flatten()\n",
    "y_pred_proba_nn = y_pred_nn\n",
    "\n",
    "print(\"Neural Network Accuracy:\", accuracy_score(y_val, (y_pred_proba_nn > 0.5).astype(int)))\n",
    "print(\"Neural Network AUC-ROC:\", roc_auc_score(y_val, y_pred_proba_nn))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "a61172db",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-12-03T09:09:52.656853Z",
     "iopub.status.busy": "2024-12-03T09:09:52.656334Z",
     "iopub.status.idle": "2024-12-03T09:09:53.254068Z",
     "shell.execute_reply": "2024-12-03T09:09:53.252962Z"
    },
    "papermill": {
     "duration": 0.937506,
     "end_time": "2024-12-03T09:09:53.256554",
     "exception": false,
     "start_time": "2024-12-03T09:09:52.319048",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model_names = ['XGBoost', 'CatBoost', 'Random Forest', 'Extra Trees', 'Neural Network']\n",
    "accuracies = [0.9525, 0.9533, 0.9493, 0.9487, 0.9467]  \n",
    "auc_roc_scores = [0.9544, 0.9526, 0.9366, 0.9260, 0.9318]  \n",
    "\n",
    "performance_df = pd.DataFrame({\n",
    "    'Model': model_names,\n",
    "    'Accuracy': accuracies,\n",
    "    'AUC-ROC': auc_roc_scores\n",
    "})\n",
    "plt.figure(figsize=(14, 8))\n",
    "sns.set(style=\"ticks\")  # Ticks style for a sharp and professional look\n",
    "\n",
    "ax1 = plt.subplot(1, 2, 1)\n",
    "sns.barplot(x='Accuracy', y='Model', data=performance_df, palette='crest', ax=ax1)\n",
    "ax1.set_title('Model Accuracy Comparison', fontsize=18, fontweight='bold', color='black')\n",
    "ax1.set_xlabel('Accuracy', fontsize=14)\n",
    "ax1.set_ylabel('Model', fontsize=14)\n",
    "ax1.set_xlim(0, 1)\n",
    "\n",
    "for p in ax1.patches:\n",
    "    ax1.annotate(f'{p.get_width():.4f}', (p.get_x() + p.get_width() - 0.02, p.get_y() + p.get_height() / 2),\n",
    "                 ha='center', va='center', fontsize=12, color='white', fontweight='bold')\n",
    "\n",
    "ax2 = plt.subplot(1, 2, 2)\n",
    "sns.barplot(x='AUC-ROC', y='Model', data=performance_df, palette='rocket', ax=ax2)\n",
    "ax2.set_title('Model AUC-ROC Comparison', fontsize=18, fontweight='bold', color='black')\n",
    "ax2.set_xlabel('AUC-ROC', fontsize=14)\n",
    "ax2.set_ylabel('Model', fontsize=14)\n",
    "ax2.set_xlim(0, 1)\n",
    "\n",
    "for p in ax2.patches:\n",
    "    ax2.annotate(f'{p.get_width():.4f}', (p.get_x() + p.get_width() - 0.02, p.get_y() + p.get_height() / 2),\n",
    "                 ha='center', va='center', fontsize=12, color='white', fontweight='bold')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ca81602d",
   "metadata": {
    "papermill": {
     "duration": 0.281874,
     "end_time": "2024-12-03T09:09:53.822770",
     "exception": false,
     "start_time": "2024-12-03T09:09:53.540896",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kaggle": {
   "accelerator": "none",
   "dataSources": [
    {
     "databundleVersionId": 9709193,
     "sourceId": 84894,
     "sourceType": "competition"
    }
   ],
   "dockerImageVersionId": 30804,
   "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": "3.10.14"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 334.246592,
   "end_time": "2024-12-03T09:09:55.832115",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2024-12-03T09:04:21.585523",
   "version": "2.6.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}