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

Download:

Current browse context:

cs.LG

Change to browse by:

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 > Machine Learning

Title: Multi-Agent Determinantal Q-Learning

Abstract: Centralized training with decentralized execution has become an important paradigm in multi-agent learning. Though practical, current methods rely on restrictive assumptions to decompose the centralized value function across agents for execution. In this paper, we eliminate this restriction by proposing multi-agent determinantal Q-learning. Our method is established on Q-DPP, an extension of determinantal point process (DPP) with partition-matroid constraint to multi-agent setting. Q-DPP promotes agents to acquire diverse behavioral models; this allows a natural factorization of the joint Q-functions with no need for \emph{a priori} structural constraints on the value function or special network architectures. We demonstrate that Q-DPP generalizes major solutions including VDN, QMIX, and QTRAN on decentralizable cooperative tasks. To efficiently draw samples from Q-DPP, we adopt an existing sample-by-projection sampler with theoretical approximation guarantee. The sampler also benefits exploration by coordinating agents to cover orthogonal directions in the state space during multi-agent training. We evaluate our algorithm on various cooperative benchmarks; its effectiveness has been demonstrated when compared with the state-of-the-art.
Comments: ICML 2020
Subjects: Machine Learning (cs.LG); Multiagent Systems (cs.MA)
Cite as: arXiv:2006.01482 [cs.LG]
  (or arXiv:2006.01482v4 [cs.LG] for this version)

Submission history

From: Yaodong Yang Mr. [view email]
[v1] Tue, 2 Jun 2020 09:32:48 GMT (5870kb,D)
[v2] Wed, 3 Jun 2020 16:18:26 GMT (5871kb,D)
[v3] Sun, 7 Jun 2020 12:43:52 GMT (6167kb,D)
[v4] Tue, 9 Jun 2020 17:50:25 GMT (6167kb,D)

Link back to: arXiv, form interface, contact.