Adversarial attacks on cooperative multi-agent deep reinforcement learning: a dynamic group-based adversarial example transferability method

Abstract Existing research shows that cooperative multi-agent deep reinforcement learning (c-MADRL) is vulnerable to adversarial attacks, and c-MADRL is increasingly being applied to safety-critical domains. However, the robustness of c-MADRL against adversarial attacks has not been fully studied. I...

Full description

Bibliographic Details

Published in:	Complex & Intelligent Systems
Main Authors:	Lixia Zan, Xiangbin Zhu, Zhao-Long Hu
Format:	Article
Language:	English
Published:	Springer 2023-07-01
Subjects:	Multi-agent reinforcement learning Adversarial attack Dynamic grouping Transfer attack Attack efficiency
Online Access:	https://doi.org/10.1007/s40747-023-01145-w