Skip to main navigation Skip to search Skip to main content

Fitness Critics for Multiagent Learning: Extended Abstract

  • Oregon State University
  • Galway-Mayo Institute of Technology

Research output: Chapter in Book or Conference Publication/Proceeding › Conference Publication › peer-review

4 Citations (Scopus)

Abstract

This paper presents fitness critics as noise-reduced fitness evaluation functions for cooperative coevolution of multiagent teams. Noise in the evaluation function can have a negative impact on evolutionary learning, especially in multiagent learning. The evaluation function for an individual agent can be noisy as each agent in the team changes their policies independently. Existing noise-reduction methods are not suitable for efficient evolutionary algorithms. The noise-reduction provided by actor-critic methods cannot be directly applied to the evolutionary algorithms when learning policies in episodic problems as the critic provides stepwise evaluations but evolutionary algorithms require episodic fitness evaluations. Fitness critics can provide noise-reduced episodic fitness evaluations by aggregating the multiple stepwise evaluations of an intermediate critic. With demonstrations on the tightly coupled multi-rover domain, this paper shows that teams trained with fitness critics achieve comparable or increased team performance scores compared to teams trained with other noise-reduction methods.

Original languageEnglish
Title of host publicationInternational Symposium on Multi-Robot and Multi-Agent Systems, MRS 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages222-224
Number of pages3
ISBN (Electronic)9781728128764
DOIs
Publication statusPublished - Aug 2019
Externally publishedYes
Event2nd International Symposium on Multi-Robot and Multi-Agent Systems, MRS 2019 - New Brunswick, United States
Duration: 22 Aug 2019 → 23 Aug 2019

Publication series

NameInternational Symposium on Multi-Robot and Multi-Agent Systems, MRS 2019

Conference

Conference2nd International Symposium on Multi-Robot and Multi-Agent Systems, MRS 2019
Country/TerritoryUnited States
CityNew Brunswick
Period22/08/19 → 23/08/19

Fingerprint

Dive into the research topics of 'Fitness Critics for Multiagent Learning: Extended Abstract'. Together they form a unique fingerprint.

Cite this