Learn the data analysis and processes behind consumer lending and how to develop, validate & monitor credit risk scorecards using analytics & SAS. This class is all about credit risk scorecard development / logistic regression model building using SAS. The entire model building and validation exercise has been explained in a step by step easy to understand manner.
Some of the discussion item would be:
- Introduction to scorecard modelling
- How do you decide the performance window?
- Exclusions and bad definition (Roll rate analysis)
- How do you perform data treatment?
- How to go for variable selection? How to deal with numeric variables and character variables?
- How do you treat multicollinearity scientifically?
- Sampling, model development plain vanilla, model selection technique, Final model selection and validation
- How do you interpret SAS output and develop next SAS code accordingly?
- Step by step workout – model development on an example data set
- Information value, WOE, Fine and coarse classing, KS statistics,Somers’D,Concordant, Brier score
- Final scorecard template
