MARL-LNS: Cooperative Multi-agent Reinforcement Learning via Large Neighborhoods Search

arXiv:2404.03101v1 Announce Type: cross
Abstract: Cooperative multi-agent reinforcement learning (MARL) has been an increasingly important research topic in the last half-decade because of its great potential for real-world applications. Because of the curse of dimensionality, the popular "centralized training decentralized execution" framework requires a long time in training, yet still cannot converge efficiently. In this paper, we propose a general training framework, MARL-LNS, to algorithmically address these issues by training on alternating subsets of agents using existing deep MARL algorithms …

abstract agent applications arxiv cs.lg cs.ma decentralized dimensionality framework multi-agent popular reinforcement reinforcement learning research search the curse of dimensionality training type via world