SMOTE.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('Data/df_transformed.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.drop(columns='ID')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\huyho\\anaconda3\\envs\\python-jupyter\\Lib\\site-packages\\seaborn\\_oldcore.py:1498: FutureWarning: is_categorical_dtype is deprecated and will be removed in a future version. Use isinstance(dtype, CategoricalDtype) instead\n",
      "  if pd.api.types.is_categorical_dtype(vector):\n",
      "c:\\Users\\huyho\\anaconda3\\envs\\python-jupyter\\Lib\\site-packages\\seaborn\\_oldcore.py:1498: FutureWarning: is_categorical_dtype is deprecated and will be removed in a future version. Use isinstance(dtype, CategoricalDtype) instead\n",
      "  if pd.api.types.is_categorical_dtype(vector):\n",
      "c:\\Users\\huyho\\anaconda3\\envs\\python-jupyter\\Lib\\site-packages\\seaborn\\_oldcore.py:1498: FutureWarning: is_categorical_dtype is deprecated and will be removed in a future version. Use isinstance(dtype, CategoricalDtype) instead\n",
      "  if pd.api.types.is_categorical_dtype(vector):\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Distribution of Non-Defaulter vs Defaulter')"
      ]
     },
     "execution_count": 98,
     "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": [
    "sns.countplot(data=df, x='Status')\n",
    "plt.title('Distribution of Non-Defaulter vs Defaulter')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Splitting the Data\n",
    "We need to splitting the data to get the original testing set before sampling\n",
    "The main goal is to fit the model either with the dataframes that were sample or the original testing set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_set, test_set = train_test_split(df, test_size=0.1, random_state=1, stratify=Target)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SMOTE\n",
    "The advantage of SMOTE is it synthesizes new examples from the minority class, as a type of data augmentation for tabular data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "from imblearn.over_sampling import SMOTE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Status\n",
       "0    100058\n",
       "1     32525\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_set['Status'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "sm = SMOTE(random_state=1)\n",
    "X = train_set.drop(columns='Status')\n",
    "y = train_set['Status']\n",
    "X_res, y_res = sm.fit_resample(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Resampled dataset shape Counter({0: 100058, 1: 100058})\n"
     ]
    }
   ],
   "source": [
    "from collections import Counter\n",
    "print('Resampled dataset shape %s' % Counter(y_res))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = X_res\n",
    "y_train = y_res\n",
    "x_test = test_set.drop(columns='Status')\n",
    "y_test = test_set['Status']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier,GradientBoostingClassifier\n",
    "from xgboost import XGBClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.metrics import precision_score,recall_score,accuracy_score,roc_auc_score,roc_curve\n",
    "from sklearn.metrics import confusion_matrix,auc,r2_score,f1_score,confusion_matrix,roc_curve,precision_recall_curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\huyho\\anaconda3\\envs\\python-jupyter\\Lib\\site-packages\\sklearn\\linear_model\\_logistic.py:469: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic Regression for test set: 0.5077382568558241\n",
      "Train Accuracy: 0.5639079333986288\n",
      "CART for test set: 1.0\n",
      "Train Accuracy: 1.0\n"
     ]
    }
   ],
   "source": [
    "#Logistic Regression and CART\n",
    "lr= LogisticRegression()\n",
    "dt= DecisionTreeClassifier(max_depth=10)\n",
    "classifiers=[('Logistic Regression',lr),('CART',dt)]\n",
    "\n",
    "for clf_name, clf in classifiers:\n",
    "    clf.fit(x_train, y_train)\n",
    "    y_pred = clf.predict(x_test)\n",
    "    print(f'{clf_name} for test set: {accuracy_score(y_test,y_pred)}')\n",
    "    print(f'Train Accuracy: {clf.score(x_train, y_train)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC SCORE:0.5886125671259248\n",
      "F1 Score: 0.39048579593208943\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC SCORE:1.0\n",
      "F1 Score: 1.0\n"
     ]
    }
   ],
   "source": [
    "for clf_name,clf in classifiers:\n",
    "    y_pred = clf.predict(x_test)\n",
    "    y_pred_proba=clf.predict_proba(x_test)[:,1]\n",
    "    fpr,tpr,thresholds=roc_curve(y_test,y_pred_proba)\n",
    "    plt.plot(fpr,tpr)\n",
    "    plt.xlabel('FPR')\n",
    "    plt.ylabel('TPR')\n",
    "    plt.title(clf_name)\n",
    "    plt.show()\n",
    "    print(f'AUC SCORE:{roc_auc_score(y_test,y_pred_proba)}')\n",
    "    print(f'F1 Score: {f1_score(y_test, y_pred)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Bagging"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import BaggingClassifier\n",
    "dt=DecisionTreeClassifier(max_depth=10,min_samples_leaf=16)\n",
    "bc=BaggingClassifier(estimator=dt,n_estimators=300,n_jobs=-1)\n",
    "bc.fit(x_train,y_train)\n",
    "y_pred=bc.predict(x_test)\n",
    "print(accuracy_score(y_test,y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# AdaBoostClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\huyho\\anaconda3\\envs\\python-jupyter\\Lib\\site-packages\\sklearn\\ensemble\\_weight_boosting.py:527: FutureWarning: The SAMME.R algorithm (the default) is deprecated and will be removed in 1.6. Use the SAMME algorithm to circumvent this warning.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import AdaBoostClassifier\n",
    "ad=AdaBoostClassifier(estimator=dt,n_estimators=1)\n",
    "ad.fit(x_train,y_train)\n",
    "y_pred=ad.predict(x_test)\n",
    "print(accuracy_score(y_test,y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_curve\n",
    "y_pred_proba=ad.predict_proba(x_test)[:,1]\n",
    "fpr,tpr,thresholds=roc_curve(y_test,y_pred_proba)\n",
    "plt.plot(fpr,tpr)\n",
    "plt.xlabel('FPR')\n",
    "plt.ylabel('TPR')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Deep Learning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\huyho\\anaconda3\\envs\\python-jupyter\\Lib\\site-packages\\keras\\src\\layers\\core\\dense.py:87: 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__(activity_regularizer=activity_regularizer, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 1ms/step - accuracy: 0.7490 - loss: 0.4905 - val_accuracy: 1.0000 - val_loss: 0.0672\n",
      "Epoch 2/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 975us/step - accuracy: 0.9785 - loss: 0.0894 - val_accuracy: 1.0000 - val_loss: 0.0064\n",
      "Epoch 3/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 910us/step - accuracy: 0.9800 - loss: 0.0677 - val_accuracy: 0.9999 - val_loss: 0.0019\n",
      "Epoch 4/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 926us/step - accuracy: 0.9797 - loss: 0.0605 - val_accuracy: 1.0000 - val_loss: 4.7617e-04\n",
      "Epoch 5/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 936us/step - accuracy: 0.9800 - loss: 0.0564 - val_accuracy: 1.0000 - val_loss: 1.9730e-04\n",
      "Epoch 6/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 923us/step - accuracy: 0.9801 - loss: 0.0538 - val_accuracy: 1.0000 - val_loss: 9.8871e-05\n",
      "Epoch 7/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 901us/step - accuracy: 0.9809 - loss: 0.0523 - val_accuracy: 1.0000 - val_loss: 5.2388e-05\n",
      "Epoch 8/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 906us/step - accuracy: 0.9797 - loss: 0.0542 - val_accuracy: 1.0000 - val_loss: 3.6587e-05\n",
      "Epoch 9/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 923us/step - accuracy: 0.9801 - loss: 0.0529 - val_accuracy: 1.0000 - val_loss: 2.6061e-05\n",
      "Epoch 10/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 909us/step - accuracy: 0.9809 - loss: 0.0522 - val_accuracy: 1.0000 - val_loss: 1.2137e-05\n",
      "Epoch 11/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 902us/step - accuracy: 0.9805 - loss: 0.0524 - val_accuracy: 1.0000 - val_loss: 1.6563e-05\n",
      "Epoch 12/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 934us/step - accuracy: 0.9813 - loss: 0.0506 - val_accuracy: 1.0000 - val_loss: 9.4393e-06\n",
      "Epoch 13/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 934us/step - accuracy: 0.9810 - loss: 0.0513 - val_accuracy: 1.0000 - val_loss: 3.6588e-06\n",
      "Epoch 14/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 919us/step - accuracy: 0.9806 - loss: 0.0522 - val_accuracy: 1.0000 - val_loss: 3.5787e-06\n",
      "Epoch 15/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 909us/step - accuracy: 0.9798 - loss: 0.0536 - val_accuracy: 1.0000 - val_loss: 1.8573e-06\n",
      "Epoch 16/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 959us/step - accuracy: 0.9806 - loss: 0.0515 - val_accuracy: 1.0000 - val_loss: 1.3752e-06\n",
      "Epoch 17/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 920us/step - accuracy: 0.9810 - loss: 0.0513 - val_accuracy: 1.0000 - val_loss: 1.0984e-06\n",
      "Epoch 18/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 955us/step - accuracy: 0.9802 - loss: 0.0537 - val_accuracy: 1.0000 - val_loss: 1.1218e-06\n",
      "Epoch 19/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 1ms/step - accuracy: 0.9806 - loss: 0.0518 - val_accuracy: 1.0000 - val_loss: 6.4649e-07\n",
      "Epoch 20/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 1ms/step - accuracy: 0.9797 - loss: 0.0536 - val_accuracy: 1.0000 - val_loss: 2.8009e-07\n",
      "Epoch 21/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m9s\u001b[0m 926us/step - accuracy: 0.9806 - loss: 0.0524 - val_accuracy: 1.0000 - val_loss: 1.0958e-04\n",
      "Epoch 22/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 946us/step - accuracy: 0.9811 - loss: 0.0510 - val_accuracy: 1.0000 - val_loss: 2.4695e-07\n",
      "Epoch 23/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 919us/step - accuracy: 0.9796 - loss: 0.0539 - val_accuracy: 1.0000 - val_loss: 2.0804e-07\n",
      "Epoch 24/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 930us/step - accuracy: 0.9808 - loss: 0.0513 - val_accuracy: 1.0000 - val_loss: 1.6877e-07\n",
      "Epoch 25/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 918us/step - accuracy: 0.9804 - loss: 0.0522 - val_accuracy: 1.0000 - val_loss: 6.4538e-08\n",
      "Epoch 26/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 951us/step - accuracy: 0.9805 - loss: 0.0518 - val_accuracy: 1.0000 - val_loss: 1.1413e-07\n",
      "Epoch 27/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 935us/step - accuracy: 0.9803 - loss: 0.0523 - val_accuracy: 1.0000 - val_loss: 1.6116e-07\n",
      "Epoch 28/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 948us/step - accuracy: 0.9812 - loss: 0.0505 - val_accuracy: 1.0000 - val_loss: 8.3207e-08\n",
      "Epoch 29/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 932us/step - accuracy: 0.9804 - loss: 0.0530 - val_accuracy: 1.0000 - val_loss: 2.4159e-08\n",
      "Epoch 30/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 916us/step - accuracy: 0.9803 - loss: 0.0523 - val_accuracy: 1.0000 - val_loss: 3.1015e-08\n",
      "Epoch 31/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 931us/step - accuracy: 0.9803 - loss: 0.0525 - val_accuracy: 1.0000 - val_loss: 1.6538e-08\n",
      "Epoch 32/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - accuracy: 0.9807 - loss: 0.0523 - val_accuracy: 1.0000 - val_loss: 4.7312e-08\n",
      "Epoch 33/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 908us/step - accuracy: 0.9804 - loss: 0.0523 - val_accuracy: 1.0000 - val_loss: 5.4483e-08\n",
      "Epoch 34/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 935us/step - accuracy: 0.9802 - loss: 0.0527 - val_accuracy: 1.0000 - val_loss: 6.0412e-08\n",
      "Epoch 35/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 916us/step - accuracy: 0.9806 - loss: 0.0520 - val_accuracy: 1.0000 - val_loss: 5.2901e-08\n",
      "Epoch 36/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 981us/step - accuracy: 0.9811 - loss: 0.0509 - val_accuracy: 1.0000 - val_loss: 3.2728e-08\n",
      "Epoch 37/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - accuracy: 0.9803 - loss: 0.0532 - val_accuracy: 1.0000 - val_loss: 1.7091e-08\n",
      "Epoch 38/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - accuracy: 0.9805 - loss: 0.0522 - val_accuracy: 1.0000 - val_loss: 1.2457e-08\n",
      "Epoch 39/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 1ms/step - accuracy: 0.9812 - loss: 0.0509 - val_accuracy: 1.0000 - val_loss: 3.4099e-09\n",
      "Epoch 40/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 1ms/step - accuracy: 0.9800 - loss: 0.0545 - val_accuracy: 1.0000 - val_loss: 1.0207e-09\n",
      "Epoch 41/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 1ms/step - accuracy: 0.9807 - loss: 0.0519 - val_accuracy: 1.0000 - val_loss: 1.3367e-07\n",
      "Epoch 42/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 1ms/step - accuracy: 0.9802 - loss: 0.0528 - val_accuracy: 1.0000 - val_loss: 2.0943e-09\n",
      "Epoch 43/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 1ms/step - accuracy: 0.9804 - loss: 0.0524 - val_accuracy: 1.0000 - val_loss: 1.2549e-09\n",
      "Epoch 44/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 1ms/step - accuracy: 0.9803 - loss: 0.0527 - val_accuracy: 1.0000 - val_loss: 1.4059e-09\n",
      "Epoch 45/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 1ms/step - accuracy: 0.9810 - loss: 0.0512 - val_accuracy: 1.0000 - val_loss: 1.5129e-09\n",
      "Epoch 46/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - accuracy: 0.9806 - loss: 0.0518 - val_accuracy: 1.0000 - val_loss: 6.2245e-10\n",
      "Epoch 47/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - accuracy: 0.9801 - loss: 0.0527 - val_accuracy: 1.0000 - val_loss: 6.6568e-10\n",
      "Epoch 48/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 924us/step - accuracy: 0.9798 - loss: 0.0535 - val_accuracy: 1.0000 - val_loss: 5.2340e-10\n",
      "Epoch 49/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - accuracy: 0.9800 - loss: 0.0532 - val_accuracy: 1.0000 - val_loss: 1.3627e-10\n",
      "Epoch 50/50\n",
      "\u001b[1m5003/5003\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 923us/step - accuracy: 0.9798 - loss: 0.0531 - val_accuracy: 1.0000 - val_loss: 1.9024e-10\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.optimizers import Adam\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from keras.layers import Dense, BatchNormalization, Dropout, LSTM\n",
    "from keras.models import Sequential\n",
    "#from keras.utils import to_categorical\n",
    "#from keras.optimizers import Adam\n",
    "from tensorflow.keras import regularizers\n",
    "from keras import callbacks\n",
    "sc = StandardScaler()\n",
    "x_train = sc.fit_transform(x_train)\n",
    "x_test = sc.transform(x_test)\n",
    "#Early stopping\n",
    "early_stopping = callbacks.EarlyStopping(\n",
    "    min_delta=0.001, # minimium amount of change to count as an improvement\n",
    "    patience=20, # how many epochs to wait before stopping\n",
    "    restore_best_weights=True,\n",
    ")\n",
    "\n",
    "# Initialising the NN\n",
    "model = Sequential()\n",
    "\n",
    "# layers\n",
    "\n",
    "model.add(Dense(units = 31, kernel_initializer = 'uniform', activation = 'relu', input_dim = 31))\n",
    "model.add(Dense(units = 31, kernel_initializer = 'uniform', activation = 'relu'))\n",
    "model.add(Dense(units = 16, kernel_initializer = 'uniform', activation = 'relu'))\n",
    "model.add(Dropout(0.25))\n",
    "model.add(Dense(units = 8, kernel_initializer = 'uniform', activation = 'relu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(units = 1, kernel_initializer = 'uniform', activation = 'sigmoid'))\n",
    "\n",
    "# Compiling the ANN\n",
    "opt = Adam(learning_rate=0.00009)\n",
    "model.compile(optimizer = opt, loss = 'binary_crossentropy', metrics = ['accuracy'])\n",
    "\n",
    "# Train the ANN\n",
    "history = model.fit(x_train, y_train, batch_size = 32, epochs = 50 , validation_split=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plotting training and validation accuracy over epochs\n",
    "\n",
    "history_df = pd.DataFrame(history.history)\n",
    "\n",
    "plt.plot(history_df.loc[:, ['accuracy']], \"#BDE2E2\", label='Training accuracy')\n",
    "plt.plot(history_df.loc[:, ['val_accuracy']], \"#C2C4E2\", label='Validation accuracy')\n",
    "\n",
    "plt.title('Training and Validation accuracy')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m461/461\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 673us/step\n"
     ]
    }
   ],
   "source": [
    "# Predicting the test set results\n",
    "y_pred = model.predict(x_test)\n",
    "y_pred = (y_pred > 0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# confusion matrix\n",
    "cmap1 = sns.diverging_palette(260,-10,s=50, l=75, n=5, as_cmap=True)\n",
    "plt.subplots(figsize=(12,8))\n",
    "cf_matrix = confusion_matrix(y_test, y_pred)\n",
    "sns.heatmap(cf_matrix/np.sum(cf_matrix), cmap = cmap1, annot = True, annot_kws = {'size':15})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      1.00      1.00     11118\n",
      "           1       1.00      1.00      1.00      3614\n",
      "\n",
      "    accuracy                           1.00     14732\n",
      "   macro avg       1.00      1.00      1.00     14732\n",
      "weighted avg       1.00      1.00      1.00     14732\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report, accuracy_score\n",
    "print(classification_report(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9999321205538962\n"
     ]
    }
   ],
   "source": [
    "print(accuracy_score(y_test,y_pred))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "python-jupyter",
   "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": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}