arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.06555cs.LGcs.MAstat.ML
收藏

随机网络系统中的多智能体强化学习

Multi-Agent Reinforcement Learning in Stochastic Networked Systems

Caltech(加州理工学院);Tsinghua University(清华大学)

Yiheng Lin, Guannan Qu, Longbo Huang, Adam Wierman

摘要:

我们研究了随机网络中的多智能体强化学习。我们的目标是找到本地化的政策,最大限度地提高(折扣)的全球回报。一般来说,可扩展性在这种设置中是一个挑战,因为全局状态/动作空间的大小可以是代理数量的指数。可扩展算法仅在依赖关系是静态的、固定的和局部的情况下是已知的,例如,在一个固定的、时不变的底层图中的邻居之间。在这项工作中,我们提出了一个可扩展的Actor Critic框架,该框架适用于依赖关系可以是非本地和随机的设置,并提供了一个有限时间的误差界,显示收敛速度如何取决于网络中信息传播的速度。此外,作为我们的分析的副产品,我们获得了新的有限时间收敛结果的一般随机逼近方案和时间差学习状态聚合,适用于网络系统中的MARL设置之外。

英文摘要:

We study multi-agent reinforcement learning (MARL) in a stochastic network of agents. The objective is to find localized policies that maximize the (discounted) global reward. In general, scalability is a challenge in this setting because the size of the global state/action space can be exponential in the number of agents. Scalable algorithms are only known in cases where dependencies are static, fixed and local, e.g., between neighbors in a fixed, time-invariant underlying graph. In this work, we propose a Scalable Actor Critic framework that applies in settings where the dependencies can be non-local and stochastic, and provide a finite-time error bound that shows how the convergence rate depends on the speed of information spread in the network. Additionally, as a byproduct of our analysis, we obtain novel finite-time convergence results for a general stochastic approximation scheme and for temporal difference learning with state aggregation, which apply beyond the setting of MARL in networked systems.

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。

↑