Credit Card Default Prediction by Usha KanchukatlaCredit Card Default Prediction by Usha Kanchukatla

Credit Card Default Prediction

Portfolio project
Usha Kanchukatla
Predictive Analytics (MAR 664) · Pace University · Mar–Apr 2026
End-to-end machine learning pipeline on 30,000 credit records in R. I compared
Logistic Regression, Decision Tree, and Random Forest to predict credit card default. Random Forest performed best, with an AUC of 0.78, and PAY_0 (repayment status) emerged as the dominant predictor.
GitHub: github.com/kanchukatlausha19-web/Predictive-Analytics-Projects

0 likes · 0 views · Nov 18, 2026

End-to-end machine learning pipeline on 30,000 credit records in R. Random Forest (AUC 0.78) outperformed Logistic Regression and Decision Tree, with PAY_0 as t

Tags