Improving Interpretability and Regularization in Deep Learning | IEEE Journals & Magazine | IEEE Xplore