Credit Card Fraud Detection

This project builds a credit card fraud detection system using machine learning on a highly imbalanced dataset. We used Random Forest for classification and SMOTE to handle data imbalance.

Dataset

The dataset contains anonymized features (V1 to V28), Time, Amount, and a Class label indicating fraud or non-fraud.

Key Steps:

1. Data Preprocessing (Scaling, Handling Imbalance).
2. Model Building with Random Forest Classifier.
3. Evaluation with accuracy, precision, recall, F1-score, and ROC-AUC.
4. SMOTE was used to balance the dataset.

Dependencies:

• Python 3.x
• Pandas, NumPy
• Scikit-learn
• Imbalanced-learn
• Matplotlib, Seaborn