arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.12010cs.LGstat.ML
收藏

QTRAN ++:改进的协作多智能体强化学习的值转换

QTRAN++: Improved Value Transformation for Cooperative Multi-Agent Reinforcement Learning

KAIST(韩国科学技术院)

Kyunghwan Son, Sungsoo Ahn, Roben Delos Reyes, Jinwoo Shin, Yung Yi

摘要:

QTRAN是一种多智能体强化学习(MARL)算法,能够学习最大类的联合作用价值函数。然而,尽管它有很强的理论保证,它在复杂环境中表现出很差的经验性能,如星际争霸多智能体挑战赛(SMAC)。在本文中,我们确定的性能瓶颈的QTRAN,并提出了一个大大改进的版本,创造QTRAN++。我们的收益来自(i)稳定QTRAN的训练目标,(ii)消除QTRAN的动作值估计器之间的严格角色分离,以及(iii)引入多头混合网络进行值转换。通过广泛的评估,我们确认我们的诊断是正确的,并且QTRAN++成功地弥合了经验性能和理论保证之间的差距。特别是,QTRAN++在SMAC环境中新实现了最先进的性能。代码将被释放。

英文摘要:

QTRAN is a multi-agent reinforcement learning (MARL) algorithm capable of learning the largest class of joint-action value functions up to date. However, despite its strong theoretical guarantee, it has shown poor empirical performance in complex environments, such as Starcraft Multi-Agent Challenge (SMAC). In this paper, we identify the performance bottleneck of QTRAN and propose a substantially improved version, coined QTRAN++. Our gains come from (i) stabilizing the training objective of QTRAN, (ii) removing the strict role separation between the action-value estimators of QTRAN, and (iii) introducing a multi-head mixing network for value transformation. Through extensive evaluation, we confirm that our diagnosis is correct, and QTRAN++ successfully bridges the gap between empirical performance and theoretical guarantee. In particular, QTRAN++ newly achieves state-of-the-art performance in the SMAC environment. The code will be released.

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。

↑