Sufficient Markov Decision Processes with Alternating Deep Neural Networks

Open in new window