Design and implement a Credit scorecard for a Credit company to mitigate Loan defaults Used two datsets - Credit bureau data and DEmographic data of the customers Merged the datasets , cleaned and imputed missing values. Used ROSE to balance the dataset as defaults were a minority (4%) Used WOE analysis to detrmine of IV of variables Checked performance of Logistic Regression and Random Trees
arunval / credit-scorecard--capstone-project Goto Github PK
View Code? Open in Web Editor NEWDesign and implement a Credit scorecard for a Credit company to mitigate Loan defaults