{ "cells": [ { "cell_type": "code", "execution_count": 5, "id": "a0ddb1e3", "metadata": {}, "outputs": [], "source": [ "import numpy as np\n", "import pandas as pd\n", "import matplotlib.pyplot as plt\n", "import seaborn as sns\n", "\n", "file=r'C:\\Users\\kkmal\\venv\\Customer_Churn_Prediction\\notebooks\\feature_transformed_telecom.csv'\n", "\n", "df=pd.read_csv(file)" ] }, { "cell_type": "code", "execution_count": 6, "id": "699e73e9", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
StreamingMoviesContractPaperlessBillingPaymentMethodMonthlyChargesTotalChargesChurnhas_partnerhas_dependentshas_phoneservice
0-1Month-to-monthYesElectronic check29.8529.85No100
1-1One yearNoMailed check56.951889.50No001
2-1Month-to-monthYesMailed check53.85108.15Yes001
3-1One yearNoBank transfer (automatic)42.301840.75No000
4-1Month-to-monthYesElectronic check70.70151.65Yes001
.................................
70381One yearYesMailed check84.801990.50No111
70391One yearYesCredit card (automatic)103.207362.90No111
7040-1Month-to-monthYesElectronic check29.60346.45No110
7041-1Month-to-monthYesMailed check74.40306.60Yes101
70421Two yearYesBank transfer (automatic)105.656844.50No001
\n", "

7043 rows × 10 columns

\n", "
" ], "text/plain": [ " StreamingMovies Contract PaperlessBilling \\\n", "0 -1 Month-to-month Yes \n", "1 -1 One year No \n", "2 -1 Month-to-month Yes \n", "3 -1 One year No \n", "4 -1 Month-to-month Yes \n", "... ... ... ... \n", "7038 1 One year Yes \n", "7039 1 One year Yes \n", "7040 -1 Month-to-month Yes \n", "7041 -1 Month-to-month Yes \n", "7042 1 Two year Yes \n", "\n", " PaymentMethod MonthlyCharges TotalCharges Churn \\\n", "0 Electronic check 29.85 29.85 No \n", "1 Mailed check 56.95 1889.50 No \n", "2 Mailed check 53.85 108.15 Yes \n", "3 Bank transfer (automatic) 42.30 1840.75 No \n", "4 Electronic check 70.70 151.65 Yes \n", "... ... ... ... ... \n", "7038 Mailed check 84.80 1990.50 No \n", "7039 Credit card (automatic) 103.20 7362.90 No \n", "7040 Electronic check 29.60 346.45 No \n", "7041 Mailed check 74.40 306.60 Yes \n", "7042 Bank transfer (automatic) 105.65 6844.50 No \n", "\n", " has_partner has_dependents has_phoneservice \n", "0 1 0 0 \n", "1 0 0 1 \n", "2 0 0 1 \n", "3 0 0 0 \n", "4 0 0 1 \n", "... ... ... ... \n", "7038 1 1 1 \n", "7039 1 1 1 \n", "7040 1 1 0 \n", "7041 1 0 1 \n", "7042 0 0 1 \n", "\n", "[7043 rows x 10 columns]" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df[['StreamingMovies', 'Contract', 'PaperlessBilling', 'PaymentMethod',\n", " 'MonthlyCharges', 'TotalCharges', 'Churn', 'has_partner',\n", " 'has_dependents', 'has_phoneservice']]" ] }, { "cell_type": "code", "execution_count": 7, "id": "d1921129", "metadata": {}, "outputs": [], "source": [ "from sklearn.preprocessing import OneHotEncoder\n", "ohe = OneHotEncoder(drop='first',handle_unknown='ignore')\n", "categorical_cols = ['Contract', 'PaperlessBilling', 'PaymentMethod']\n", "encoded_data = ohe.fit_transform(df[categorical_cols]).toarray()\n", "encoded_df = pd.DataFrame(encoded_data, columns=ohe.get_feature_names_out(categorical_cols))\n", "df = pd.concat([df.drop(columns=categorical_cols), encoded_df], axis=1,)\n", "\n", "df=df.drop(columns=['Unnamed: 0'])" ] }, { "cell_type": "code", "execution_count": 8, "id": "79f55bab", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "RangeIndex: 7043 entries, 0 to 7042\n", "Data columns (total 28 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 gender 7043 non-null int64 \n", " 1 SeniorCitizen 7043 non-null int64 \n", " 2 tenure 7043 non-null int64 \n", " 3 OnlineSecurity 7043 non-null int64 \n", " 4 OnlineBackup 7043 non-null int64 \n", " 5 DeviceProtection 7043 non-null int64 \n", " 6 TechSupport 7043 non-null int64 \n", " 7 StreamingTV 7043 non-null int64 \n", " 8 StreamingMovies 7043 non-null int64 \n", " 9 MonthlyCharges 7043 non-null float64\n", " 10 TotalCharges 7043 non-null float64\n", " 11 Churn 7043 non-null str \n", " 12 has_partner 7043 non-null int64 \n", " 13 has_dependents 7043 non-null int64 \n", " 14 has_phoneservice 7043 non-null int64 \n", " 15 has_multiplelines 7043 non-null int64 \n", " 16 internet_service_encoded 7043 non-null int64 \n", " 17 is_automatic 7043 non-null int64 \n", " 18 Product_Count 7043 non-null int64 \n", " 19 Is_High_Risk_Integration 7043 non-null int64 \n", " 20 Is_Fully_Integrated 7043 non-null int64 \n", " 21 Mod_Security 7043 non-null int64 \n", " 22 Contract_One year 7043 non-null float64\n", " 23 Contract_Two year 7043 non-null float64\n", " 24 PaperlessBilling_Yes 7043 non-null float64\n", " 25 PaymentMethod_Credit card (automatic) 7043 non-null float64\n", " 26 PaymentMethod_Electronic check 7043 non-null float64\n", " 27 PaymentMethod_Mailed check 7043 non-null float64\n", "dtypes: float64(8), int64(19), str(1)\n", "memory usage: 1.5 MB\n" ] } ], "source": [ "df.info()" ] }, { "cell_type": "code", "execution_count": null, "id": "5ecf91f5", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": 9, "id": "bcc7b0e0", "metadata": {}, "outputs": [], "source": [ "from sklearn.model_selection import train_test_split\n", "X = df.drop(columns=['Churn'])\n", "y=df['Churn'].map({'No': 0, 'Yes': 1})\n", "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25, random_state=42,stratify=y)" ] }, { "cell_type": "code", "execution_count": 10, "id": "fed9684b", "metadata": {}, "outputs": [], "source": [ "from sklearn.preprocessing import StandardScaler\n", "scaler = StandardScaler()\n", "numerical_cols = ['MonthlyCharges', 'TotalCharges']\n", "X_train[numerical_cols] = scaler.fit_transform(X_train[numerical_cols])\n", "X_test[numerical_cols] = scaler.transform(X_test[numerical_cols])\n" ] }, { "cell_type": "code", "execution_count": 11, "id": "99deca84", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "((5282, 27), (5282,), (1761, 27), (1761,))" ] }, "execution_count": 11, "metadata": {}, "output_type": "execute_result" } ], "source": [ "X_train.shape, y_train.shape, X_test.shape, y_test.shape" ] }, { "cell_type": "code", "execution_count": 12, "id": "34a5a236", "metadata": {}, "outputs": [], "source": [ "from sklearn.linear_model import LogisticRegression, RidgeClassifier\n", "from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier, IsolationForest\n", "from sklearn.tree import DecisionTreeClassifier\n", "from xgboost import XGBClassifier\n", "from sklearn.svm import SVC" ] }, { "cell_type": "code", "execution_count": 13, "id": "10a147c0", "metadata": {}, "outputs": [], "source": [ "import numpy as np\n", "# Assuming y_train is a pandas Series or numpy array\n", "# Calculate scale_pos_weight for XGBoost: (count of negative examples / count of positive examples)\n", "ratio = float(np.sum(y_train == 0) / np.sum(y_train == 1))\n", "\n", "models_configuration = {\n", " \"Logistic Regression\": {\n", " \"model\": LogisticRegression(max_iter=1000, class_weight=\"balanced\"),\n", " \"params\": {\n", " \"C\": [0.1, 1.0, 10.0],\n", " \"solver\": [\"liblinear\", \"lbfgs\"]\n", " }\n", " },\n", " \"Ridge Classifier\": {\n", " \"model\": RidgeClassifier(class_weight=\"balanced\"),\n", " \"params\": {\n", " \"alpha\": [0.1, 1.0, 10.0]\n", " }\n", " },\n", " \"Decision Tree\": {\n", " \"model\": DecisionTreeClassifier(class_weight=\"balanced\"),\n", " \"params\": {\n", " \"max_depth\": [None, 5, 10, 20],\n", " \"min_samples_split\": [2, 5, 10]\n", " }\n", " },\n", " \"Random Forest\": {\n", " \"model\": RandomForestClassifier(class_weight=\"balanced\"),\n", " \"params\": {\n", " \"n_estimators\": [50, 100, 200],\n", " \"max_depth\": [None, 10, 20]\n", " }\n", " },\n", " \"Gradient Boosting\": {\n", " # GradientBoostingClassifier doesn't have class_weight, \n", " # so we rely on subsampling or tuning, OR use HistGradientBoostingClassifier which does.\n", " \"model\": GradientBoostingClassifier(), \n", " \"params\": {\n", " \"learning_rate\": [0.01, 0.1, 0.2],\n", " \"n_estimators\": [50, 100, 200]\n", " }\n", " },\n", " \"XGBoost\": {\n", " # scale_pos_weight controls the balance of positive/negative weights\n", " \"model\": XGBClassifier(eval_metric=\"logloss\", scale_pos_weight=ratio),\n", " \"params\": {\n", " \"learning_rate\": [0.01, 0.1, 0.2],\n", " \"max_depth\": [3, 5, 7],\n", " \"n_estimators\": [50, 100, 200]\n", " }\n", " },\n", " \"SVC\": {\n", " \"model\": SVC(class_weight=\"balanced\"),\n", " \"params\": {\n", " \"C\": [0.1, 1.0, 10.0],\n", " \"kernel\": [\"rbf\", \"linear\"]\n", " }\n", " }\n", "}" ] }, { "cell_type": "code", "execution_count": 14, "id": "c3d40bb9", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Tuning Logistic Regression...\n", "Best Score: 0.8440\n", "\n", "Tuning Ridge Classifier...\n", "Best Score: 0.8419\n", "\n", "Tuning Decision Tree...\n", "Best Score: 0.8157\n", "\n", "Tuning Random Forest...\n", "Best Score: 0.8413\n", "\n", "Tuning Gradient Boosting...\n", "Best Score: 0.8457\n", "\n", "Tuning XGBoost...\n", "Best Score: 0.8458\n", "\n", "Tuning SVC...\n", "Best Score: 0.8425\n", "\n" ] } ], "source": [ "from imblearn.over_sampling import SMOTE\n", "\n", "# Apply SMOTE only to your TRAINING data (never the test data!)\n", "smote = SMOTE(random_state=42)\n", "X_train_balanced, y_train_balanced = smote.fit_resample(X_train, y_train)\n", "\n", "# Now, you can use your original loop from the previous answer, \n", "# just pass X_train_balanced and y_train_balanced into clf.fit()\n", "\n", "from sklearn.model_selection import GridSearchCV\n", "\n", "\n", "best_models = {}\n", "\n", "for model_name, config in models_configuration.items():\n", " print(f\"Tuning {model_name}...\")\n", "\n", " clf = GridSearchCV(config[\"model\"], config[\"params\"], cv=5, scoring=\"roc_auc\", n_jobs=-1)\n", " clf.fit(X_train_balanced, y_train_balanced)\n", " \n", " # Save the best model and score\n", " best_models[model_name] = {\n", " \"best_model\": clf.best_estimator_,\n", " \"best_score\": clf.best_score_,\n", " \"best_params\": clf.best_params_\n", " }\n", " \n", " print(f\"Best Score: {clf.best_score_:.4f}\\n\")" ] }, { "cell_type": "code", "execution_count": 15, "id": "b5b27faf", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "{'best_model': XGBClassifier(base_score=None, booster=None, callbacks=None,\n", " colsample_bylevel=None, colsample_bynode=None,\n", " colsample_bytree=None, device=None, early_stopping_rounds=None,\n", " enable_categorical=False, eval_metric='logloss',\n", " feature_types=None, feature_weights=None, gamma=None,\n", " grow_policy=None, importance_type=None,\n", " interaction_constraints=None, learning_rate=0.1, max_bin=None,\n", " max_cat_threshold=None, max_cat_to_onehot=None,\n", " max_delta_step=None, max_depth=3, max_leaves=None,\n", " min_child_weight=None, missing=nan, monotone_constraints=None,\n", " multi_strategy=None, n_estimators=50, n_jobs=None,\n", " num_parallel_tree=None, ...),\n", " 'best_score': np.float64(0.8458143320981766),\n", " 'best_params': {'learning_rate': 0.1, 'max_depth': 3, 'n_estimators': 50}}" ] }, "execution_count": 15, "metadata": {}, "output_type": "execute_result" } ], "source": [ "best_models['XGBoost']" ] }, { "cell_type": "code", "execution_count": null, "id": "2b6fc460", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": 23, "id": "49e312ab", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Classification Report:\n", " precision recall f1-score support\n", "\n", " 0 0.94 0.61 0.74 1294\n", " 1 0.45 0.88 0.60 467\n", "\n", " accuracy 0.68 1761\n", " macro avg 0.69 0.75 0.67 1761\n", "weighted avg 0.81 0.68 0.70 1761\n", "\n" ] } ], "source": [ "ratio = float(np.sum(y_train == 0) / np.sum(y_train == 1))\n", "\n", "xgc=XGBClassifier(eval_metric=\"logloss\", scale_pos_weight=ratio,learning_rate=0.1, max_depth=3, n_estimators=50)\n", "xgc.fit(X_train_balanced, y_train_balanced)\n", "y_pred_proba=xgc.predict_proba(X_test)\n", "y_pred=xgc.predict(X_test)\n", "from sklearn.metrics import classification_report, confusion_matrix, roc_auc_score, roc_curve\n", "print(\"Classification Report:\")\n", "print(classification_report(y_test, y_pred))" ] }, { "cell_type": "code", "execution_count": 24, "id": "36d02a5c", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import matplotlib.pyplot as plt\n", "from sklearn.metrics import RocCurveDisplay, PrecisionRecallDisplay\n", "\n", "# Set up a 1x2 plotting area\n", "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(14, 6))\n", "\n", "# 1. Plot ROC Curve on the first axis\n", "RocCurveDisplay.from_estimator(xgc, X_test, y_test, ax=ax1)\n", "ax1.plot([0, 1], [0, 1], 'k--', label='Random (AUC = 0.50)')\n", "ax1.set_title('ROC Curve (Can look overly optimistic)')\n", "ax1.grid(True)\n", "\n", "# 2. Plot Precision-Recall Curve on the second axis\n", "PrecisionRecallDisplay.from_estimator(xgc, X_test, y_test, ax=ax2)\n", "# Baseline for PR curve is the percentage of positive cases (467 / 1761)\n", "baseline = 467 / 1761\n", "ax2.plot([0, 1], [baseline, baseline], 'k--', label=f'Random (AP = {baseline:.2f})')\n", "ax2.set_title('Precision-Recall Curve (The Truth for Churn)')\n", "ax2.grid(True)\n", "\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": null, "id": "c93784e1", "metadata": {}, "outputs": [], "source": [ "import pickle\n", "\n", "pickle.dump(xgc,open(r'C:\\Users\\kkmal\\venv\\Customer_Churn_Prediction\\models\\xgboost_model.pkl','wb'))\n", "pickle.dump(scaler,open(r'C:\\Users\\kkmal\\venv\\Customer_Churn_Prediction\\models\\scaler.pkl','wb'))\n", "pickle.dump(ohe,open(r'C:\\Users\\kkmal\\venv\\Customer_Churn_Prediction\\models\\ohe.pkl','wb'))" ] }, { "cell_type": "code", "execution_count": null, "id": "e4dc184e", "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "ML-Default", "language": "python", "name": "ml_env" } }, "nbformat": 4, "nbformat_minor": 5 }