Fix_1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "410aa378-65f6-4f49-9927-249e8a8825c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "import time\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.preprocessing import RobustScaler\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV, cross_validate\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.tree import DecisionTreeClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "aaa0b840-2f1b-4567-b9a1-80e9cd2b450e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load data\n",
    "df = pd.read_csv('data/creditcard.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "31855945-c63a-4573-b424-e3f58c92b9e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scaling\n",
    "scaler = RobustScaler()\n",
    "\n",
    "df['scaled_amount'] = scaler.fit_transform(df['Amount'].values.reshape(-1, 1))\n",
    "df['scaled_time'] = scaler.fit_transform(df['Time'].values.reshape(-1, 1))\n",
    "\n",
    "df.drop(['Amount', 'Time'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "97adbcac-4cf3-4041-b3ba-309db03246d5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split train and test\n",
    "train, test = train_test_split(df, test_size=0.2, shuffle=True, random_state=42, stratify=df['Class'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "3cfa6ab7-61a9-468b-a35e-db2e9553adcc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Check test dataset class distribution\n",
    "test_val_cnt = test['Class'].value_counts().reset_index()\n",
    "plt.bar(test_val_cnt['index'], test_val_cnt['Class'], color=['red', 'blue'], tick_label=['normal', 'fraud'])\n",
    "plt.title('Class Distribution of test dataset')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "0a1e29b2-12b1-4a3d-8341-abd19a0a28b2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Length of fraud data: 394\n",
      "Length of normal data: 1576\n"
     ]
    }
   ],
   "source": [
    "# Under Sampling train dataset\n",
    "fraud = train[train['Class']==1].reset_index(drop=True)\n",
    "normal = train[train['Class']==0]\n",
    "\n",
    "num_of_normal = len(fraud) * 4    # 2:8 비율\n",
    "\n",
    "normal = normal.sample(n=num_of_normal, random_state=42, ignore_index=True)\n",
    "\n",
    "print(f'Length of fraud data: {len(fraud)}')\n",
    "print(f'Length of normal data: {len(normal)}')\n",
    "\n",
    "train = pd.concat([fraud, normal], ignore_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "d669682f-355f-4170-94d7-afd0b9bddb61",
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Delete outliers\n",
    "# def del_outliers(df, col):\n",
    "#     fraud = df[df['Class']==1][col].values\n",
    "#     q25, q75 = np.percentile(fraud, 25), np.percentile(fraud, 75)\n",
    "#     iqr = q75 - q25\n",
    "#     cut_off = iqr * 1.5\n",
    "#     lower, upper = q25 - cut_off, q75 + cut_off\n",
    "#     df.drop(df[(df[col] > upper) | (df[col] < lower)].index, inplace=True)\n",
    "#     print(f'Number of data after {col} ouliers removal: {len(df)}')\n",
    "    \n",
    "# columns = ['V10', 'V12', 'V14', 'V17']\n",
    "\n",
    "# for col in columns:\n",
    "#     del_outliers(train, col)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "ad267c57-a1da-408b-806b-648ee48ddf58",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "< Logistic Regression Results >\n",
      "Total fit time: 0.1902923107147217\n",
      "The best params: {'C': 0.01, 'penalty': 'l2'}\n",
      "Test accuracy: 97.69%\n",
      "\n",
      "\n",
      "< K-Neighbors Results >\n",
      "Total fit time: 0.02692546844482422\n",
      "The best params: {'algorithm': 'auto', 'n_neighbors': 4}\n",
      "Test accuracy: 95.92%\n",
      "\n",
      "\n",
      "< SVC Results >\n",
      "Total fit time: 0.2857597827911377\n",
      "The best params: {'C': 1, 'kernel': 'rbf'}\n",
      "Test accuracy: 97.6%\n",
      "\n",
      "\n",
      "< Decision Tree Results >\n",
      "Total fit time: 2.0971799373626707\n",
      "The best params: {'criterion': 'entropy', 'max_depth': 4, 'min_samples_leaf': 8}\n",
      "Test accuracy: 95.05%\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Train Classifier\n",
    "X_train = train.drop('Class', axis=1)\n",
    "y_train = train['Class']\n",
    "X_test = test.drop('Class', axis=1)\n",
    "y_test = test['Class']\n",
    "\n",
    "model = LogisticRegression(max_iter=10000)\n",
    "params = {'penalty':['l2'], 'C':[0.001, 0.01, 0.1, 1, 10, 100, 1000]}\n",
    "grid = GridSearchCV(model, params, scoring=['accuracy', 'roc_auc'], refit='roc_auc', return_train_score=True, cv=5)\n",
    "grid.fit(X_train, y_train)\n",
    "\n",
    "print('< Logistic Regression Results >')\n",
    "print(f'Total fit time: {grid.cv_results_[\"mean_fit_time\"].sum()}')\n",
    "print(f'The best params: {grid.best_params_}')\n",
    "print(f'Test accuracy: {round(grid.score(X_test, y_test) * 100, 2)}%')\n",
    "print('\\n')\n",
    "\n",
    "model = KNeighborsClassifier()\n",
    "params = {\"n_neighbors\": list(range(2,5,1)), 'algorithm': ['auto', 'ball_tree', 'kd_tree', 'brute']}\n",
    "grid = GridSearchCV(model, params, scoring=['accuracy', 'roc_auc'], refit='roc_auc', return_train_score=True, cv=5)\n",
    "grid.fit(X_train, y_train)\n",
    "\n",
    "print('< K-Neighbors Results >')\n",
    "print(f'Total fit time: {grid.cv_results_[\"mean_fit_time\"].sum()}')\n",
    "print(f'The best params: {grid.best_params_}')\n",
    "print(f'Test accuracy: {round(grid.score(X_test, y_test) * 100, 2)}%')\n",
    "print('\\n')\n",
    "\n",
    "model = SVC()\n",
    "params = {'C': [0.5, 0.7, 0.9, 1], 'kernel': ['rbf', 'poly', 'sigmoid', 'linear']}\n",
    "grid = GridSearchCV(model, params, scoring=['accuracy', 'roc_auc'], refit='roc_auc', return_train_score=True, cv=5)\n",
    "grid.fit(X_train, y_train)\n",
    "\n",
    "print('< SVC Results >')\n",
    "print(f'Total fit time: {grid.cv_results_[\"mean_fit_time\"].sum()}')\n",
    "print(f'The best params: {grid.best_params_}')\n",
    "print(f'Test accuracy: {round(grid.score(X_test, y_test) * 100, 2)}%')\n",
    "print('\\n')\n",
    "\n",
    "model = DecisionTreeClassifier()\n",
    "params = {\n",
    "    \"criterion\": [\"gini\", \"entropy\"], \n",
    "    \"max_depth\": list(range(2,10,1)), \n",
    "    \"min_samples_leaf\": list(range(2,10,1))\n",
    "}\n",
    "grid = GridSearchCV(model, params, scoring=['accuracy', 'roc_auc'], refit='roc_auc', return_train_score=True, cv=5)\n",
    "grid.fit(X_train, y_train)\n",
    "\n",
    "print('< Decision Tree Results >')\n",
    "print(f'Total fit time: {grid.cv_results_[\"mean_fit_time\"].sum()}')\n",
    "print(f'The best params: {grid.best_params_}')\n",
    "print(f'Test accuracy: {round(grid.score(X_test, y_test) * 100, 2)}%')\n",
    "print('\\n')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.8.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}