Mean-field learning for satisfactory solutions

Handle URI:
http://hdl.handle.net/10754/575818
Title:
Mean-field learning for satisfactory solutions
Authors:
Tembine, Hamidou; Tempone, Raul ( 0000-0003-1967-4446 ) ; Vilanova, Pedro ( 0000-0001-6620-6261 )
Abstract:
One of the fundamental challenges in distributed interactive systems is to design efficient, accurate, and fair solutions. In such systems, a satisfactory solution is an innovative approach that aims to provide all players with a satisfactory payoff anytime and anywhere. In this paper we study fully distributed learning schemes for satisfactory solutions in games with continuous action space. Considering games where the payoff function depends only on own-action and an aggregate term, we show that the complexity of learning systems can be significantly reduced, leading to the so-called mean-field learning. We provide sufficient conditions for convergence to a satisfactory solution and we give explicit convergence time bounds. Then, several acceleration techniques are used in order to improve the convergence rate. We illustrate numerically the proposed mean-field learning schemes for quality-of-service management in communication networks. © 2013 IEEE.
KAUST Department:
Center for Uncertainty Quantification in Computational Science and Engineering (SRI-UQ); Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division; Applied Mathematics and Computational Science Program; Stochastic Numerics Research Group
Publisher:
Institute of Electrical and Electronics Engineers (IEEE)
Journal:
52nd IEEE Conference on Decision and Control
Conference/Event name:
52nd IEEE Conference on Decision and Control, CDC 2013
Issue Date:
Dec-2013
DOI:
10.1109/CDC.2013.6760653
ARXIV:
arXiv:1210.4657v1
Type:
Conference Paper
ISSN:
01912216
ISBN:
9781467357173
Appears in Collections:
Conference Papers; Applied Mathematics and Computational Science Program; Applied Mathematics and Computational Science Program; Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division; Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division

Full metadata record

DC FieldValue Language
dc.contributor.authorTembine, Hamidouen
dc.contributor.authorTempone, Raulen
dc.contributor.authorVilanova, Pedroen
dc.date.accessioned2015-08-24T09:27:01Zen
dc.date.available2015-08-24T09:27:01Zen
dc.date.issued2013-12en
dc.identifier.isbn9781467357173en
dc.identifier.issn01912216en
dc.identifier.doi10.1109/CDC.2013.6760653en
dc.identifier.urihttp://hdl.handle.net/10754/575818en
dc.description.abstractOne of the fundamental challenges in distributed interactive systems is to design efficient, accurate, and fair solutions. In such systems, a satisfactory solution is an innovative approach that aims to provide all players with a satisfactory payoff anytime and anywhere. In this paper we study fully distributed learning schemes for satisfactory solutions in games with continuous action space. Considering games where the payoff function depends only on own-action and an aggregate term, we show that the complexity of learning systems can be significantly reduced, leading to the so-called mean-field learning. We provide sufficient conditions for convergence to a satisfactory solution and we give explicit convergence time bounds. Then, several acceleration techniques are used in order to improve the convergence rate. We illustrate numerically the proposed mean-field learning schemes for quality-of-service management in communication networks. © 2013 IEEE.en
dc.publisherInstitute of Electrical and Electronics Engineers (IEEE)en
dc.titleMean-field learning for satisfactory solutionsen
dc.typeConference Paperen
dc.contributor.departmentCenter for Uncertainty Quantification in Computational Science and Engineering (SRI-UQ)en
dc.contributor.departmentComputer, Electrical and Mathematical Sciences and Engineering (CEMSE) Divisionen
dc.contributor.departmentApplied Mathematics and Computational Science Programen
dc.contributor.departmentStochastic Numerics Research Groupen
dc.identifier.journal52nd IEEE Conference on Decision and Controlen
dc.conference.date10 December 2013 through 13 December 2013en
dc.conference.name52nd IEEE Conference on Decision and Control, CDC 2013en
dc.conference.locationFlorenceen
dc.identifier.arxividarXiv:1210.4657v1en
kaust.authorTembine, Hamidouen
kaust.authorTempone, Raulen
kaust.authorVilanova, Pedroen
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.