Online Program

Return to main conference page

All Times EDT

Friday, June 5
Machine Learning
Interpretable and Fair Machine Learning in Finance
Fri, Jun 5, 11:15 AM - 12:50 PM
TBD
 

A New Approach to Providing Explanations for Machine Learning Algorithms (308291)

*Tom Prendergast, Synchrony Financial 

The Transparent Explanations Algorithm (TEA) provides the ability to generate top-n influential variables with a rank for each prediction of a machine learning model. It is a generic capability intended to be used for any machine learning model, black-box or otherwise. Originally conceptualized with the intent to explain outcomes of credit models for purposes of adverse action, it can also function as a robust model diagnostic solution for nearly any predictive model.