An RBF based neuro-dynamic approach for the control of stochastic dynamic systems

Patrinos, PK; Sarimveis, H

dc.contributor.author	Patrinos, PK	en
dc.contributor.author	Sarimveis, H	en
dc.date.accessioned	2014-03-01T02:49:59Z
dc.date.available	2014-03-01T02:49:59Z
dc.date.issued	2005	en
dc.identifier.issn	14746670	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/34856
dc.relation.uri	http://www.scopus.com/inward/record.url?eid=2-s2.0-79960740645&partnerID=40&md5=ccdec55739708756eadadfcd9d67027e	en
dc.subject	Markov decision processes	en
dc.subject	Optimal control	en
dc.subject	Radial base function networks	en
dc.subject	Uncertain dynamic systems	en
dc.subject.other	Bellman equations	en
dc.subject.other	Compact representation	en
dc.subject.other	Management problems	en
dc.subject.other	Markov Decision Processes	en
dc.subject.other	Neuro dynamic programming	en
dc.subject.other	Optimal controls	en
dc.subject.other	Policy evaluation	en
dc.subject.other	Policy iteration	en
dc.subject.other	Radial base function	en
dc.subject.other	Stochastic dynamic systems	en
dc.subject.other	Uncertain dynamic systems	en
dc.subject.other	Automation	en
dc.subject.other	Control	en
dc.subject.other	Dynamic programming	en
dc.subject.other	Inventory control	en
dc.subject.other	Learning algorithms	en
dc.subject.other	Markov processes	en
dc.subject.other	Radial basis function networks	en
dc.subject.other	Process control	en
dc.title	An RBF based neuro-dynamic approach for the control of stochastic dynamic systems	en
heal.type	conferenceItem	en
heal.publicationDate	2005	en
heal.abstract	This paper presents a neuro-dynamic programming methodology for the control of markov decision processes. The proposed method can be considered as a variant of the optimistic policy iteration, where radial basis function (RBF) networks are employed as a compact representation of the cost-to-go function and the λ-LSPE is used for policy evaluation. We also emphasize the reformulation of the Bellman equation around the post-decision state in order to circumvent the calculation of the expectation. The proposed algorithm is applied to a retailer-inventory management problem. Copyright © 2005 IFAC.	en
heal.journalName	IFAC Proceedings Volumes (IFAC-PapersOnline)	en
dc.identifier.volume	16	en
dc.identifier.spage	52	en
dc.identifier.epage	57	en