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

Download:

Current browse context:

cs.AI

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 > Artificial Intelligence

Title: The Power of Communication in a Distributed Multi-Agent System

Abstract: Single-Agent (SA) Reinforcement Learning systems have shown outstanding re-sults on non-stationary problems. However, Multi-Agent Reinforcement Learning(MARL) can surpass SA systems generally and when scaling. Furthermore, MAsystems can be super-powered by collaboration, which can happen through ob-serving others, or a communication system used to share information betweencollaborators. Here, we developed a distributed MA learning mechanism withthe ability to communicate based on decentralised partially observable Markovdecision processes (Dec-POMDPs) and Graph Neural Networks (GNNs). Minimis-ing the time and energy consumed by training Machine Learning models whileimproving performance can be achieved by collaborative MA mechanisms. Wedemonstrate this in a real-world scenario, an offshore wind farm, including a set ofdistributed wind turbines, where the objective is to maximise collective efficiency.Compared to a SA system, MA collaboration has shown significantly reducedtraining time and higher cumulative rewards in unseen and scaled scenarios.
Comments: Cooperative AI Workshop at the 35th Conference on Neural Information Processing Systems (NeurIPS 2021), Sydney, Australia
Subjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
Cite as: arXiv:2111.15611 [cs.AI]
  (or arXiv:2111.15611v3 [cs.AI] for this version)

Submission history

From: Philipp Siedler [view email]
[v1] Tue, 30 Nov 2021 18:00:58 GMT (17022kb,D)
[v2] Wed, 1 Dec 2021 11:41:13 GMT (10404kb,D)
[v3] Tue, 14 Dec 2021 21:11:56 GMT (10404kb,D)

Link back to: arXiv, form interface, contact.