A Regularized Maximum Figure-of-Merit (rMFoM) Approach to Supervised and Semi-Supervised Learning | IEEE Journals & Magazine | IEEE Xplore