Building prediction models with the right balance between performance and interpretability is currently a great challenge in machine learning. A large number of recent studies have focused on either building intrinsically ...