'''python
Repository for CODSOFT Data Science Internship tasks import pandas as pd from sklearn.model_selection import train_test_split from sklearn.preprocessing import StandardScaler from sklearn.linear_model import LogisticRegression from sklearn.ensemble import RandomForestClassifier from sklearn.metrics import classification_report, confusion_matrix, roc_auc_score, roc_curve import matplotlib.pyplot as plt import seaborn as sns from imblearn.over_sampling import SMOTE # For handling class imbalance from imblearn.under_sampling import RandomUnderSampler import numpy as np
Ensure the path is correct, the dataset should be in the same directory, or specify the correct path
try: df = pd.read_csv('creditcard.csv') except FileNotFoundError: print("Error: creditcard.csv not found. Please make sure the file is in the correct directory.") exit()
print("Dataset shape:", df.shape) print("Dataset head:") print(df.head()) print("Dataset info:") df.info()
print("Missing values before cleaning:") print(df.isnull().sum().sum())
Handle missing values (if any) - using 0 for simplicity. You might want to use mean, median, or imputation.
df.fillna(0, inplace=True) # Replace NaN values with 0
print("Missing values after cleaning:") print(df.isnull().sum().sum())
print("Duplicate values before cleaning:", df.duplicated().sum())
df = df.drop_duplicates()
print("Duplicate values after cleaning:", df.duplicated().sum())
print("Class distribution:") print(df['Class'].value_counts())
X = df.drop('Class', axis=1) y = df['Class']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)
scaler = StandardScaler() X_train_scaled = scaler.fit_transform(X_train) X_test_scaled = scaler.transform(X_test)
smote = SMOTE(random_state=42) X_train_smote, y_train_smote = smote.fit_resample(X_train_scaled, y_train)
rus = RandomUnderSampler(random_state=42) X_train_rus, y_train_rus = rus.fit_resample(X_train_scaled, y_train)
Apply SMOTE first to oversample the minority class, then apply RandomUnderSampler to the majority class
smote_combined = SMOTE(sampling_strategy=0.5, random_state=42) # You can adjust sampling_strategy rus_combined = RandomUnderSampler(sampling_strategy=0.8, random_state=42) # You can adjust sampling_strategy X_train_combined, y_train_combined = smote_combined.fit_resample(X_train_scaled, y_train) # Apply SMOTE X_train_combined, y_train_combined = rus_combined.fit_resample(X_train_combined, y_train_combined) # Apply RUS
def evaluate_model(model, X_test, y_test, model_name="Model"): """ Evaluates a classification model and prints performance metrics. Also generates a ROC curve.
Args:
model: Trained classification model.
X_test: Scaled test features.
y_test: True labels for the test set.
model_name: Name of the model for display purposes.
"""
y_pred = model.predict(X_test)
y_pred_proba = model.predict_proba(X_test)[:, 1] # Probabilities for the positive class
print(f"\n--- {model_name} ---")
print("Confusion Matrix:")
print(confusion_matrix(y_test, y_pred))
print("Classification Report:")
print(classification_report(y_test, y_pred))
print("ROC AUC Score:", roc_auc_score(y_test, y_pred_proba))
# Generate ROC Curve
fpr, tpr, thresholds = roc_curve(y_test, y_pred_proba)
plt.figure(figsize=(8, 6))
plt.plot(fpr, tpr, color='darkorange', lw=2, label='ROC curve')
plt.plot([0, 1], [0, 1], color='navy', lw=2, linestyle='--')
plt.xlabel('False Positive Rate')
plt.ylabel('True Positive Rate')
plt.title(f'ROC Curve for {model_name}')
plt.legend(loc="lower right")
plt.show()
return y_pred, y_pred_proba # Return both predictions and probabilities
logistic_model = LogisticRegression(random_state=42, solver='liblinear') #solver specified logistic_model.fit(X_train_scaled, y_train) #default y_pred_logistic, y_pred_proba_logistic = evaluate_model(logistic_model, X_test_scaled, y_test, "Logistic Regression (Original)")
logistic_model_smote = LogisticRegression(random_state=42, solver='liblinear') logistic_model_smote.fit(X_train_smote, y_train_smote) y_pred_logistic_smote, y_pred_proba_logistic_smote = evaluate_model(logistic_model_smote, X_test_scaled, y_test, "Logistic Regression (SMOTE)")
logistic_model_rus = LogisticRegression(random_state=42, solver='liblinear') logistic_model_rus.fit(X_train_rus, y_train_rus) y_pred_logistic_rus, y_pred_proba_logistic_rus = evaluate_model(logistic_model_rus, X_test_scaled, y_test, "Logistic Regression (RUS)")
logistic_model_combined = LogisticRegression(random_state=42, solver='liblinear') logistic_model_combined.fit(X_train_combined, y_train_combined) y_pred_logistic_combined, y_pred_proba_logistic_combined = evaluate_model(logistic_model_combined, X_test_scaled, y_test, "Logistic Regression (Combined)")
rf_model = RandomForestClassifier(random_state=42) rf_model.fit(X_train_scaled, y_train) #default y_pred_rf, y_pred_proba_rf = evaluate_model(rf_model, X_test_scaled, y_test, "Random Forest (Original)")
rf_model_smote = RandomForestClassifier(random_state=42) rf_model_smote.fit(X_train_smote, y_train_smote) y_pred_rf_smote, y_pred_proba_rf_smote = evaluate_model(rf_model, X_test_scaled, y_test, "Random Forest (SMOTE)")
rf_model_rus = RandomForestClassifier(random_state=42) rf_model_rus.fit(X_train_rus, y_train_rus) y_pred_rf_rus, y_pred_proba_rf_rus = evaluate_model(rf_model, X_test_scaled, y_test, "Random Forest (RUS)")
rf_model_combined = RandomForestClassifier(random_state=42) rf_model_combined.fit(X_train_combined, y_train_combined) y_pred_rf_combined, y_pred_proba_rf_combined = evaluate_model(rf_model, X_test_scaled, y_test, "Random Forest (Combined)")
print("\n--- Model Comparison (ROC AUC Scores) ---") print(f"Logistic Regression (Original): {roc_auc_score(y_test, y_pred_proba_logistic):.4f}") print(f"Logistic Regression (SMOTE): {roc_auc_score(y_test, y_pred_proba_logistic_smote):.4f}") print(f"Logistic Regression (RUS): {roc_auc_score(y_test, y_pred_proba_logistic_rus):.4f}") print(f"Logistic Regression (Combined): {roc_auc_score(y_test, y_pred_proba_logistic_combined):.4f}") print(f"Random Forest (Original): {roc_auc_score(y_test, y_pred_proba_rf):.4f}") print(f"Random Forest (SMOTE): {roc_auc_score(y_test, y_pred_proba_rf_smote):.4f}") print(f"Random Forest (RUS): {roc_auc_score(y_test, y_pred_proba_rf_rus):.4f}") print(f"Random Forest (Combined): {roc_auc_score(y_test, y_pred_proba_rf_combined):.4f}")
if 'rf_model' in locals(): # Check if the variable exists feature_importances = rf_model.feature_importances_ feature_names = X.columns sorted_indices = np.argsort(feature_importances)[::-1] # Sort in descending order
plt.figure(figsize=(10, 6))
plt.title("Feature Importance (Random Forest)")
plt.bar(range(X.shape[1]), feature_importances[sorted_indices], align="center")
plt.xticks(range(X.shape[1]), feature_names[sorted_indices], rotation='vertical')
plt.xlim([-1, X.shape[1]])
plt.tight_layout()
plt.show()
from sklearn.metrics import precision_recall_curve
if 'y_pred_proba_logistic_smote' in locals(): precision, recall, thresholds = precision_recall_curve(y_test, y_pred_proba_logistic_smote)
plt.figure(figsize=(8, 6))
plt.plot(recall, precision, color='b', lw=2, label='Precision-Recall curve')
plt.xlabel('Recall')
plt.ylabel('Precision')
plt.title('Precision-Recall Curve (Logistic Regression - SMOTE)')
plt.legend(loc="lower left")
plt.show()
'''
Result:
Dataset shape: (284807, 31) Dataset head: Time V1 V2 V3 V4 V5 ... V25 V26 V27 V28 Amount Class 0 0.0 -1.359807 -0.072781 2.536347 1.378155 -0.338321 ... 0.128539 -0.189115 0.133558 -0.021053 149.62 0 1 0.0 1.191857 0.266151 0.166480 0.448154 0.060018 ... 0.167170 0.125895 -0.008983 0.014724 2.69 0 2 1.0 -1.358354 -1.340163 1.773209 0.379780 -0.503198 ... -0.327642 -0.139097 -0.055353 -0.059752 378.66 0 3 1.0 -0.966272 -0.185226 1.792993 -0.863291 -0.010309 ... 0.647376 -0.221929 0.062723 0.061458 123.50 0 4 2.0 -1.158233 0.877737 1.548718 0.403034 -0.407193 ... -0.206010 0.502292 0.219422 0.215153 69.99 0
[5 rows x 31 columns] Dataset info: <class 'pandas.core.frame.DataFrame'> RangeIndex: 284807 entries, 0 to 284806 Data columns (total 31 columns):
0 Time 284807 non-null float64 1 V1 284807 non-null float64 2 V2 284807 non-null float64 3 V3 284807 non-null float64 4 V4 284807 non-null float64 5 V5 284807 non-null float64 6 V6 284807 non-null float64 7 V7 284807 non-null float64 8 V8 284807 non-null float64 9 V9 284807 non-null float64 10 V10 284807 non-null float64 11 V11 284807 non-null float64 12 V12 284807 non-null float64 13 V13 284807 non-null float64 14 V14 284807 non-null float64 15 V15 284807 non-null float64 16 V16 284807 non-null float64 17 V17 284807 non-null float64 18 V18 284807 non-null float64 19 V19 284807 non-null float64 20 V20 284807 non-null float64 21 V21 284807 non-null float64 22 V22 284807 non-null float64 23 V23 284807 non-null float64 24 V24 284807 non-null float64 25 V25 284807 non-null float64 26 V26 284807 non-null float64 27 V27 284807 non-null float64 28 V28 284807 non-null float64 29 Amount 284807 non-null float64 30 Class 284807 non-null int64 dtypes: float64(30), int64(1) memory usage: 67.4 MB Missing values before cleaning: 0 Missing values after cleaning: 0 Duplicate values before cleaning: 1081 Duplicate values after cleaning: 0 Class distribution: Class 0 283253 1 473 Name: count, dtype: int64
--- Logistic Regression (Original) --- Confusion Matrix: [[84973 11] [ 60 74]] Classification Report: precision recall f1-score support
0 1.00 1.00 1.00 84984
1 0.87 0.55 0.68 134
accuracy 1.00 85118
macro avg 0.93 0.78 0.84 85118 weighted avg 1.00 1.00 1.00 85118
ROC AUC Score: 0.9695674058400459
--- Logistic Regression (SMOTE) --- Confusion Matrix: [[82923 2061] [ 14 120]] Classification Report: precision recall f1-score support
0 1.00 0.98 0.99 84984
1 0.06 0.90 0.10 134
accuracy 0.98 85118
macro avg 0.53 0.94 0.55 85118 weighted avg 1.00 0.98 0.99 85118
ROC AUC Score: 0.9771626897986767
--- Logistic Regression (RUS) --- Confusion Matrix: [[81639 3345] [ 15 119]] Classification Report: precision recall f1-score support
0 1.00 0.96 0.98 84984
1 0.03 0.89 0.07 134
accuracy 0.96 85118
macro avg 0.52 0.92 0.52 85118 weighted avg 1.00 0.96 0.98 85118
ROC AUC Score: 0.9736323940169246
--- Logistic Regression (Combined) --- Confusion Matrix: [[83375 1609] [ 15 119]] Classification Report: precision recall f1-score support
0 1.00 0.98 0.99 84984
1 0.07 0.89 0.13 134
accuracy 0.98 85118
macro avg 0.53 0.93 0.56 85118 weighted avg 1.00 0.98 0.99 85118
ROC AUC Score: 0.977112460853035
