arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2312.09009cs.AI
收藏

多智能体强化学习的自适应参数共享

Adaptive parameter sharing for multi-agent reinforcement learning

Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所);School of Artificial Intelligence, University of Chinese Academy of Sciences(中国科学院大学人工智能学院)

Dapeng Li, Na Lou, Bin Zhang, Zhiwei Xu, Guoliang Fan

AI总结:

针对多智能体系统中朴素参数共享难以生成差异化策略的问题,提出一种受生物学启发的方法,将不同身份智能体映射到共享网络的不同区域形成子网络,在不增加参数的前提下提升策略多样性并取得更优性能。

摘要:

参数共享作为多智能体系统中的重要技术,能有效解决大规模智能体问题中的可扩展性问题。然而,参数共享的有效性很大程度上取决于环境设置。当智能体具有不同的身份或任务时,朴素的参数共享难以生成足够差异化的策略。受生物学大脑研究的启发,我们提出了一种新颖的参数共享方法。该方法根据身份将每种类型的智能体映射到共享网络的不同区域,形成不同的子网络。因此,我们的方法能在不引入额外训练参数的情况下增加不同智能体间策略的多样性。通过在多个环境中进行的实验,我们的方法表现出优于其他参数共享方法的性能。

英文摘要:

Parameter sharing, as an important technique in multi-agent systems, can effectively solve the scalability issue in large-scale agent problems. However, the effectiveness of parameter sharing largely depends on the environment setting. When agents have different identities or tasks, naive parameter sharing makes it difficult to generate sufficiently differentiated strategies for agents. Inspired by research pertaining to the brain in biology, we propose a novel parameter sharing method. It maps each type of agent to different regions within a shared network based on their identity, resulting in distinct subnetworks. Therefore, our method can increase the diversity of strategies among different agents without introducing additional training parameters. Through experiments conducted in multiple environments, our method has shown better performance than other parameter sharing methods.

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。

↑