arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.09585cs.GTcs.LGstat.ML
收藏

基于最佳响应的多智能体强化学习策略评估与搜索

Policy Evaluation and Seeking for Multi-Agent Reinforcement Learning via Best Response

Tsinghua University(清华大学);UC Santa Barbara(加州大学圣巴巴拉分校)

Rui Yan, Xiaoming Duan, Zongying Shi, Yisheng Zhong, Jason R. Marden, Francesco Bullo

摘要:

本文介绍了两个指标(基于周期和基于内存的指标),基于一个动态博弈论的解决方案的概念,称为汇平衡,在多智能体学习的政策的评估,排名和计算。我们采用严格最佳反应动力学(SBRD)模型自私的行为在元级别的多智能体强化学习。我们的方法可以处理动态周期性行为(与基于纳什均衡和Elo评级的方法不同),并且与依赖于弱更好响应的alpha秩相比,与单智能体强化学习更兼容。我们首先考虑最大和第二大基础度量之间的差异具有已知下限的设置。有了这些知识,我们提出了一类扰动SBRD具有以下性质:只有最大度量的政策观察到非零概率的一类广泛的随机游戏有限的记忆。然后,我们考虑设置的差异的下限是未知的。对于这种设置,我们提出了一类扰动SBRD,使非零概率观察到的政策的度量不同于任何给定的公差的最佳。拟议的扰动SBRD解决了被动诱导的非平稳性,通过固定的学习代理的其他人的策略,并使用经验博弈论分析,估计由于扰动获得的每个策略配置文件的回报。

英文摘要:

This paper introduces two metrics (cycle-based and memory-based metrics), grounded on a dynamical game-theoretic solution concept called sink equilibrium, for the evaluation, ranking, and computation of policies in multi-agent learning. We adopt strict best response dynamics (SBRD) to model selfish behaviors at a meta-level for multi-agent reinforcement learning. Our approach can deal with dynamical cyclical behaviors (unlike approaches based on Nash equilibria and Elo ratings), and is more compatible with single-agent reinforcement learning than alpha-rank which relies on weakly better responses. We first consider settings where the difference between largest and second largest underlying metric has a known lower bound. With this knowledge we propose a class of perturbed SBRD with the following property: only policies with maximum metric are observed with nonzero probability for a broad class of stochastic games with finite memory. We then consider settings where the lower bound for the difference is unknown. For this setting, we propose a class of perturbed SBRD such that the metrics of the policies observed with nonzero probability differ from the optimal by any given tolerance. The proposed perturbed SBRD addresses the opponent-induced non-stationarity by fixing the strategies of others for the learning agent, and uses empirical game-theoretic analysis to estimate payoffs for each strategy profile obtained due to the perturbation.

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。

↑