ICML 2020poster215 citations

Deep Coordination Graphs

Wendelin Boehmer, Vitaly Kurin, Shimon Whiteson

Abstract

This paper introduces the deep coordination graph (DCG) for collaborative multi-agent reinforcement learning. DCG strikes a flexible trade-off between representational capacity and generalization by factoring the joint value function of all agents according to a coordination graph into payoffs between pairs of agents. The value can be maximized by local message passing along the graph, which allows training of the value function end-to-end with Q-learning. Payoff functions are approximated with deep neural networks that employ parameter sharing and low-rank approximations to significantly improve sample efficiency. We show that DCG can solve predator-prey tasks that highlight the relative overgeneralization pathology, as well as challenging StarCraft II micromanagement tasks.

BibTeX
@InProceedings{pmlr-v119-boehmer20a,
  title = 	 {Deep Coordination Graphs},
  author =       {Boehmer, Wendelin and Kurin, Vitaly and Whiteson, Shimon},
  booktitle = 	 {Proceedings of the 37th International Conference on Machine Learning},
  pages = 	 {980--991},
  year = 	 {2020},
  editor = 	 {III, Hal Daumé and Singh, Aarti},
  volume = 	 {119},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {13--18 Jul},
  publisher =    {PMLR},
  pdf = 	 {http://proceedings.mlr.press/v119/boehmer20a/boehmer20a.pdf},
  url = 	 {https://proceedings.mlr.press/v119/boehmer20a.html},
  abstract = 	 {This paper introduces the deep coordination graph (DCG) for collaborative multi-agent reinforcement learning. DCG strikes a flexible trade-off between representational capacity and generalization by factoring the joint value function of all agents according to a coordination graph into payoffs between pairs of agents. The value can be maximized by local message passing along the graph, which allows training of the value function end-to-end with Q-learning. Payoff functions are approximated with deep neural networks that employ parameter sharing and low-rank approximations to significantly improve sample efficiency. We show that DCG can solve predator-prey tasks that highlight the relative overgeneralization pathology, as well as challenging StarCraft II micromanagement tasks.}
}
Deep Coordination Graphs · ICML 2020