DM$^2$: Distributed multi-agent reinforcement learning via distribution matching
Published in Preprint, 2022
DM$^2$ allows a team of agents to perform cooperative tasks by independently imitating corresponding experts agents from a team of experts.
Recommended citation: Caroline Wang, Ishan Durugkar, Elad Liebman, Peter Stone (2022). "DM$^2$: Distributed multi-agent reinforcement learning via distribution matching" arXiv preprint.