*/Computer Science

Inspired by a reference paper on loan default prediction, I am developing an improved model that focuses not only on accuracy but also on fairness and explainability. My goal is to measure how fair the model is across different groups, calculate the probability of repayment, understand the key reasons behind repayment or default, and even predict when repayment is most likely. I also explore whether non-financial data (like education or behavioral patterns) can enhance prediction. Finally, I analyze how changes in factors such as age, education, or sex affect repayment probability, while ensuring the model remains unbiased.