We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.MA

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Multiagent Systems

Title: Directionality Reinforcement Learning to Operate Multi-Agent System without Communication

Abstract: This paper establishes directionality reinforcement learning (DRL) technique to propose the complete decentralized multi-agent reinforcement learning method which can achieve cooperation based on each agent's learning: no communication and no observation. Concretely, DRL adds the direction "agents have to learn to reach the farthest goal among reachable ones" to learning agents to operate the agents cooperatively. Furthermore, to investigate the effectiveness of the DRL, this paper compare Q-learning agent with DRL with previous learning agent in maze problems. Experimental results derive that (1) DRL performs better than the previous method in terms of the spending time, (2) the direction makes agents learn yielding action for others, and (3) DRL suggests achieving multiagent learning with few costs for any number of agents.
Comments: 5 pages, 2 figures, accepted at AAMAS Workshop on Multiagent Optimization and Learning 2020 (OptLearnMAS 2020)
Subjects: Multiagent Systems (cs.MA)
MSC classes: 68T42
ACM classes: I.2; I.2.11
Cite as: arXiv:2110.05773 [cs.MA]
  (or arXiv:2110.05773v1 [cs.MA] for this version)

Submission history

From: Fumito Uwano Dr. [view email]
[v1] Tue, 12 Oct 2021 07:05:56 GMT (15424kb)

Link back to: arXiv, form interface, contact.