Capstone Project – Home Loan Default Prediction
Course: PGPDSBA Capstone Project (Oct 2024)
This project builds a machine learning model to predict home loan defaults, aimed at assisting a bank’s consumer credit department in improving the decision-making process for home equity lines of credit. Following the Equal Credit Opportunity Act’s guidelines, the goal is to create an empirically derived, statistically sound model for credit scoring based on recent applicant data. After thorough data preprocessing, several models were tested, with the Random Forest model on oversampled data showing the best performance by prioritizing recall and reducing bias in loan approvals. Key features influencing defaults were analyzed to guide effective business recommendations.
Deliverables (for each, Milestone 1, Milestone 2, Live Presentation and Final Presentation):
- Business Report and Presentation
- Python Notebook
Github: PGPDSBA Capstone Project – Home Loan Default Prediction
