Interpretability is a very key aspect of 2nd Order’s solution to provide financial institutions with explanations to customers for certain financial decisions, such as credit card application denials as mandated under US law. In this project, we evaluated and compared the two most popular explainable prediction algorithms commonly used in the industry, Gradient Boosting Models (GBM) and Explainable Boosting Machines (EBM), on the same dataset for metrics like training and prediction time, as well as their AUC score and explainability. Our findings show that EBM performs better, has better explainability. It allows for case-by-case feature importance and can identify minor class cases more effectively than GBM, making it a better choice for credit risk assessment that meets regulatory requirements.
