This project implements a fraud detection system using a combination of unsupervised learning (Isolation Forest) for anomaly detection and supervised learning (XGBoost) for classification. The script includes data preprocessing, anomaly filtering, model training, evaluation, and visualization.
The script requires the following Python packages:
numpypandasscikit-learnxgboostmatplotlibseaborn
pip install numpy pandas scikit-learn xgboost matplotlib seabornpython fraud_detection.py