Cognition-Driven Multiagent Policy Learning Framework for Promoting Cooperation

doi:10.1109/TG.2022.3186386

CORC > 自动化研究所 > 中国科学院自动化研究所 > 复杂系统认知与决策实验室

	Cognition-Driven Multiagent Policy Learning Framework for Promoting Cooperation
	Pu, Zhiqiang 1,3; Wang, Huimu 2; Liu, Boyin 1,3; Yi, Jianqiang 1,3
刊名	IEEE TRANSACTIONS ON GAMES
	2023-09-01
卷号	15 期号:3 页码:388-398
关键词	Cognition difference coupling cognition network (CCN) deep reinforcement learning (DRL) graph convolutional network multiagent systems (MASs)
ISSN号	2475-1502
DOI	10.1109/TG.2022.3186386
通讯作者	Pu, Zhiqiang(zhiqiang.pu@ia.ac.cn)
英文摘要	Many attempts have been made to promote cooperation for multiagent systems. However, several issues that draw less attentions but may dramatically degrade the cooperation performance still exist, such as redundant information interactions among neighbors, and difficulties in understanding complex and dynamic environments from high-level cognition. To address these limitations, a cognition-driven multiagent policy (CDMAP) learning framework is proposed in this article. It includes a cognition difference network (CDN), a coupling cognition network (CCN), and a policy optimization network (PON). CDN is designed based on a variational autoencoder, where a concept of cognition difference is defined to prune redundant interactions among agents for more efficient communication. Based on the pruned topology, CCN captures the hidden representations of the surrounding environment. Several coupling graph attention layers are incorporated in CCN, each layer with different but coupling adjacent matrices, yielding a comprehensive state understanding from multiple representation spaces. Based on the captured hidden states, PON generates the final policies, where QMIX is adopted as a value factorization method to alleviate the credit-assignment problem. At last, CDMAP is evaluated through two representative multiagent games including Google Research Football and StarCraft II. The results demonstrate its superior effectiveness compared with existing methods.
资助项目	National Key Research Development Program of China ; Strategic Priority Research Program of Chinese Academy of Sciences[2020AAA0103404] ; External Cooperation Key Project of Chinese Academy Sciences[XDA27030204] ; Science and Technology Development Fund of Macau[173211KYSB20200002] ; [0025/2019/AKP]
WOS关键词	GAME
WOS研究方向	Computer Science
语种	英语
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号	WOS:001068900800007
资助机构	National Key Research Development Program of China ; Strategic Priority Research Program of Chinese Academy of Sciences ; External Cooperation Key Project of Chinese Academy Sciences ; Science and Technology Development Fund of Macau
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/53059]
专题	复杂系统认知与决策实验室
通讯作者	Pu, Zhiqiang
作者单位	1.Chinese Acad Sci, Inst Automat, Beijing 100190, Peoples R China 2.JD COM, Beijing 100176, Peoples R China 3.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 100049, Peoples R China
推荐引用方式 GB/T 7714	Pu, Zhiqiang,Wang, Huimu,Liu, Boyin,et al. Cognition-Driven Multiagent Policy Learning Framework for Promoting Cooperation[J]. IEEE TRANSACTIONS ON GAMES,2023,15(3):388-398.
APA	Pu, Zhiqiang,Wang, Huimu,Liu, Boyin,&Yi, Jianqiang.(2023).Cognition-Driven Multiagent Policy Learning Framework for Promoting Cooperation.IEEE TRANSACTIONS ON GAMES,15(3),388-398.
MLA	Pu, Zhiqiang,et al."Cognition-Driven Multiagent Policy Learning Framework for Promoting Cooperation".IEEE TRANSACTIONS ON GAMES 15.3(2023):388-398.