Tractability of interpretability via selection of group-sparse models | IEEE Conference Publication | IEEE Xplore