Money mule account detection using LightGBM + XGBoost ensemble (0.985 AUC-ROC). 125 features, 12 fraud patterns, SHAP explainability. RBIH x IIT Delhi TRYST 2025.
python machine-learning hackathon eda fintech xgboost lightgbm feature-engineering fraud-detection iit-delhi anti-money-laundering shap mule-detection rbih
-
Updated
Mar 17, 2026 - HTML