Heuristically-accelerated reinforcement learning: A comparative analysis of performance

dc.contributor.authorMARTINS, M. F.
dc.contributor.authorReinaldo Bianchi
dc.contributor.authorOrcidhttps://orcid.org/0000-0001-9097-827X
dc.date.accessioned2022-01-12T22:01:08Z
dc.date.available2022-01-12T22:01:08Z
dc.date.issued2014
dc.description.abstractThis paper presents a comparative analysis of three Reinforcement Learning algorithms (Q-learning, Q(λ)-learning and QS-learning) and their heuristically-accelerated variants (HAQL, HAQ(λ) and HAQS) where heuristics bias action selection, thus speeding up the learning. The experiments were performed in a simulated robot soccer environment which reproduces the conditions of a real competition league environment. The results clearly demonstrate that the use of heuristics substantially improves the performance of the learning algorithms. © 2014 Springer-Verlag.
dc.description.firstpage15
dc.description.lastpage27
dc.description.volume8069 LNAI
dc.identifier.citationMARTINS, M. F.; BIANCHI, R. Heuristically-accelerated reinforcement learning: A comparative analysis of performance. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), p. 15-27, 2014.
dc.identifier.doi10.1007/978-3-662-43645-5_2
dc.identifier.issn1611-3349
dc.identifier.urihttps://repositorio.fei.edu.br/handle/FEI/4047
dc.relation.ispartofLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
dc.rightsAcesso Restrito
dc.subject.otherlanguageHeuristics
dc.subject.otherlanguageReinforcement learning
dc.subject.otherlanguageRobot soccer
dc.titleHeuristically-accelerated reinforcement learning: A comparative analysis of performance
dc.typeArtigo de evento
fei.scopus.citations2
fei.scopus.eid2-s2.0-84904201099
fei.scopus.subjectAction selection
fei.scopus.subjectComparative analysis
fei.scopus.subjectHeuristics
fei.scopus.subjectQ-learning
fei.scopus.subjectRobot soccer
fei.scopus.subjectSimulated robot
fei.scopus.updated2024-05-01
fei.scopus.urlhttps://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84904201099&origin=inward
Arquivos
Coleções