Building an Interpretable Recommender via Loss-Preserving Transformation

Amit Dhurandhar, Sechan Oh, Marek Petrik

We propose a method for building an interpretable recommender system for personalizing online content and promotions. Historical data available for the system consists of customer…