A reinforcement learning based algorithm for Markov decision processes | IEEE Conference Publication | IEEE Xplore