Multi-agent reinforcement learning: an approach based on the other agent's internal model | IEEE Conference Publication | IEEE Xplore