VGN: Value Decomposition With Graph Attention Networks for Multiagent Reinforcement Learning

doi:10.1109/TNNLS.2022.3172572

CASIA OpenIR > 多模态人工智能系统全国重点实验室 > 复杂系统智能机理与平行控制团队

	VGN: Value Decomposition With Graph Attention Networks for Multiagent Reinforcement Learning
	Wei, Qinglai1,2,3 ; Li, Yugu 1; Zhang, Jie1 ; Wang, Fei-Yue1,3,4
发表期刊	IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
ISSN	2162-237X
	2022-05-18
页码	14
通讯作者	Zhang, Jie(jie.zhang@ia.ac.cn)
摘要	Although value decomposition networks and the follow on value-based studies factorizes the joint reward function to individual reward functions for a kind of cooperative multiagent reinforcement problem, in which each agent has its local observation and shares a joint reward signal, most of the previous efforts, however, ignored the graphical information between agents. In this article, a new value decomposition with graph attention network (VGN) method is developed to solve the value functions by introducing the dynamical relationships between agents. It is pointed out that the decomposition factor of an agent in our approach can be influenced by the reward signals of all the related agents and two graphical neural network-based algorithms (VGN-Linear and VGN-Nonlinear) are designed to solve the value functions of each agent. It can be proved theoretically that the present methods satisfy the factorizable condition in the centralized training process. The performance of the present methods is evaluated on the StarCraft Multiagent Challenge (SMAC) benchmark. Experiment results show that our method outperforms the state-of-the-art value-based multiagent reinforcement algorithms, especially when the tasks are with very hard level and challenging for existing methods.
关键词	Mathematical models Task analysis Games Q-learning Neural networks Behavioral sciences Training Deep learning graph attention networks (GATs) multiagent systems reinforcement learning
DOI	10.1109/TNNLS.2022.3172572
收录类别	SCI
语种	英语
资助项目	National Key Research and Development Program of China[2021YFE0206100] ; National Natural Science Foundation of China[62073321] ; National Defense Basic Scientific Research Program[JCKY2019203C029] ; Science and Technology Development Fund, Macau[0015/2020/AMJ]
项目资助者	National Key Research and Development Program of China ; National Natural Science Foundation of China ; National Defense Basic Scientific Research Program ; Science and Technology Development Fund, Macau
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Artificial Intelligence ; Computer Science, Hardware & Architecture ; Computer Science, Theory & Methods ; Engineering, Electrical & Electronic
WOS记录号	WOS:000798352100001
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计	被引频次：6[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/49411
专题	多模态人工智能系统全国重点实验室_复杂系统智能机理与平行控制团队多模态人工智能系统全国重点实验室_平行智能技术与系统团队
通讯作者	Zhang, Jie
作者单位	1.Chinese Acad Sci, Inst Automat, State Key Lab Management & Control Complex Syst, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 100049, Peoples R China 3.Macau Univ Sci & Technol, Inst Syst Engn, Macau 999078, Peoples R China 4.Qingdao Acad Intelligent Ind, Qingdao 266109, Peoples R China
第一作者单位	中国科学院自动化研究所
通讯作者单位	中国科学院自动化研究所
推荐引用方式 GB/T 7714	Wei, Qinglai,Li, Yugu,Zhang, Jie,et al. VGN: Value Decomposition With Graph Attention Networks for Multiagent Reinforcement Learning[J]. IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS,2022:14.
APA	Wei, Qinglai,Li, Yugu,Zhang, Jie,&Wang, Fei-Yue.(2022).VGN: Value Decomposition With Graph Attention Networks for Multiagent Reinforcement Learning.IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS,14.
MLA	Wei, Qinglai,et al."VGN: Value Decomposition With Graph Attention Networks for Multiagent Reinforcement Learning".IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS (2022):14.