General objective The overarching objective of this project is to develop a comprehensive Credit Scoring Model that accurately assesses the creditworthiness of potential borrowers for the Buy-Now-Pay-Later (BNPL) service offered by Bati Bank. This model will facilitate informed lending decisions by categorizing users into risk categories, predicting loan performance, and optimizing loan parameters. Specific objectives 1. Defines a proxy variable that can be used to categorize users as high risk (bad) or low risk (good) 2. Select observable features that are good predictors (have high correlation) of the default variable defined in 1) 3. Develop a model that assigns risk probability for a new customer 4. Develop a model that assigns credit score from risk probability estimates 5. Develop a model that predicts the optimal amount and duration of the loan
Figure: Comparison of different classifier models for Prediction of credit score