Distance-Penalized Active Learning via Markov Decision Processes | IEEE Conference Publication | IEEE Xplore