Training Deep Models to be Explained with Fewer Examples

Tomoharu Iwata, Yuya Yoshikawa

Although deep models achieve high predictive performance, it is difficult for humans to understand the predictions they made. Explainability is important for real-world application…