Xin-Ge Miao

Domain-Driven Feature Engineering and Multi-Machine Learning Models for Credit Card Default Prediction

This work proves the value of financial domain knowledge in feature construction and provides a stable, reproducible benchmark scheme for credit risk assessment.