arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1907.03053cs.LGcs.MAstat.ML
收藏

分布式强化学习中通信高效的多Agent Actor-Critic算法

A Communication-Efficient Multi-Agent Actor-Critic Algorithm for Distributed Reinforcement Learning

Stony Brook University(石溪大学);University of Illinois at Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校);Princeton University(普林斯顿大学);Northwestern University(西北大学)

Yixuan Lin, Kaiqing Zhang, Zhuoran Yang, Zhaoran Wang, Tamer Başar, Romeil Sandhu, Ji Liu

摘要:

本文考虑了一个分布式强化学习问题,其中多个代理的网络旨在通过仅与本地邻居通信来合作最大化全局平均收益。针对由有向图描述的可能单向通信关系,提出了一种随机通信有效的多Agent Actor-Critic算法。结果表明,该算法可以解决强连通图的问题,通过允许每个代理,只有两个标量值变量在同一时间。

英文摘要:

This paper considers a distributed reinforcement learning problem in which a network of multiple agents aim to cooperatively maximize the globally averaged return through communication with only local neighbors. A randomized communication-efficient multi-agent actor-critic algorithm is proposed for possibly unidirectional communication relationships depicted by a directed graph. It is shown that the algorithm can solve the problem for strongly connected graphs by allowing each agent to transmit only two scalar-valued variables at one time.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑