A unified framework for linear function approximation of value functions in stochastic control

e-Archivo Repository

Show simple item record

dc.contributor.author Sánchez-Fernández, Matilde
dc.contributor.author Valcárcel, Sergio
dc.contributor.author Zazo, Santiago
dc.date.accessioned 2015-06-26T11:52:21Z
dc.date.available 2015-06-26T11:52:21Z
dc.date.issued 2013-09
dc.identifier.bibliographicCitation Proceedings of the 21st European Signal Processing Conference (EUSIPCO) (2013) pp. 1-5
dc.identifier.uri http://hdl.handle.net/10016/21205
dc.description The proceeding at:21st European Signal Processing Conference (EUSIPCO 2013), took place 2013, September 9-13, in Marrakech (marroc).
dc.description.abstract This paper contributes with a unified formulation that merges previous analysis on the prediction of the performance (value function) of certain sequence of actions (policy) when an agent operates a Markov decision process with large state-space. When the states are represented by features and the value function is linearly approximated, our analysis reveals a new relationship between two common cost functions used to obtain the optimal approximation. In addition, this analysis allows us to propose an efficient adaptive algorithm that provides an unbiased linear estimate. The performance of the proposed algorithm is illustrated by simulation, showing competitive results when compared with the state-of-the-art solutions.
dc.description.sponsorship This work has been partly funded by the Spanish Ministry of Science and Innovation with the project GRE3N (TEC 2011-29006-C03-01/02/03) and in the program CONSOLIDER-INGENIO 2010 under project COMONSENS (CSD 2008-00010). This work was supported in part by the Spanish Ministry of Science and Innovation under the grants TEC2009-14219-C03-01,TEC2010-21217-C02- 02-CR4HFDVL and in the program CONSOLIDER-INGENIO 2010 under the grant CSD2008-00010 COMONSENS; and by the European Commission under the grant FP7-ICT-2009-4-248894-WHERE-2.
dc.format.extent 5
dc.format.mimetype application/pdf
dc.language.iso eng
dc.publisher IEEE - The Institute of Electrical and Electronics Engineers, Inc
dc.rights © 2013 IEEE
dc.subject.other Function approximation
dc.subject.other Markov processes
dc.subject.other Signal processing
dc.subject.other Stochastic systems
dc.title A unified framework for linear function approximation of value functions in stochastic control
dc.type bookPart
dc.type conferenceObject
dc.description.status Publicado
dc.relation.publisherversion http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6811729
dc.subject.eciencia Telecomunicaciones
dc.rights.accessRights openAccess
dc.relation.projectID Gobierno de España. TEC2011-29006-C03-02
dc.relation.projectID Gobierno de España. TEC2011-29006-C03-03
dc.type.version acceptedVersion
dc.relation.eventdate 2013, September 9-13
dc.relation.eventnumber 21
dc.relation.eventplace Marrakech (Marroc)
dc.relation.eventtitle European Signal Processing Conference (EUSIPCO 2013)
dc.relation.eventtype proceeding
dc.identifier.publicationfirstpage 1
dc.identifier.publicationlastpage 5
dc.identifier.publicationtitle Proceedings of the 21st European Signal Processing Conference (EUSIPCO)
dc.identifier.uxxi CC/0000022069
 Find Full text

Files in this item

*Click on file's image for preview. (Embargoed files's preview is not supported)


This item appears in the following Collection(s)

Show simple item record