Responsible for running as many as 20 statistical tests as part of the validation process for logistic regression credit scorecard models. Key components of this task include modifying and developing SAS Macros to perform key statistical tests which evaluate model accuracy, discriminatory power, and sensitivity to changes in model parameters.
Worked with developers to understand complex methodologies and data manipulations such as the creation and replication of pseudo default datasets used for scorecard modeling.
Developed challenger models with alternative inputs and data manipulations to provide an effective challenge to models submitted by developers.
Performed quarterly ongoing monitoring for 10 credit scorecard models, and documented results using Latex for typesetting.