Structural relational inference actor-critic for multi-agent reinforcement learning

•A novel MARL algorithm involving the interaction relationship between agents.•This algorithm is based on centralized training and decentralized execution.•We apply the variational autoencoder model to obtain the latent relations.•We employ graph attention network to gather the information of neighb...

Full description

Bibliographic Details

Published in:	Neurocomputing (Amsterdam) Vol. 459; pp. 383 - 394
Main Authors:	Zhang, Xianjie, Liu, Yu, Xu, Xiujuan, Huang, Qiong, Mao, Hangyu, Carie, Anil
Format:	Journal Article
Language:	English
Published:	Elsevier B.V 12.10.2021
Subjects:	Actor-critic Deep reinforcement learning Graph neural network Multi-agent systems Variational autoencoder Variational autoencoder Deep reinforcement learning Actor-critic Graph neural network Multi-agent systems
ISSN:	0925-2312, 1872-8286
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!