Bayesian model order selection for nonlinear system function expansions

Mitsis, GD; Jbabdi, S

dc.contributor.author	Mitsis, GD	en
dc.contributor.author	Jbabdi, S	en
dc.date.accessioned	2014-03-01T02:45:10Z
dc.date.available	2014-03-01T02:45:10Z
dc.date.issued	2008	en
dc.identifier.issn	1557170X	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/32175
dc.subject	bayesian framework	en
dc.subject	bayesian model	en
dc.subject	Exponential Decay	en
dc.subject	laguerre function	en
dc.subject	Model Order Selection	en
dc.subject	Nonlinear System	en
dc.subject	Nonlinear System Identification	en
dc.subject	Order Selection	en
dc.subject	Posterior Distribution	en
dc.subject	Prediction Error	en
dc.subject	System Identification	en
dc.subject.other	Bayesian networks	en
dc.subject.other	Expansion	en
dc.subject.other	Neural networks	en
dc.subject.other	Nonlinear systems	en
dc.subject.other	Bayesian frameworks	en
dc.subject.other	Bayesian models	en
dc.subject.other	Exponential decays	en
dc.subject.other	Free parameters	en
dc.subject.other	Laguerre	en
dc.subject.other	Laguerre basis	en
dc.subject.other	Laguerre expansion techniques	en
dc.subject.other	Laguerre functions	en
dc.subject.other	Model predictions	en
dc.subject.other	Model-order selections	en
dc.subject.other	Non linearities	en
dc.subject.other	Nonlinear systems identifications	en
dc.subject.other	Orthonormal basis	en
dc.subject.other	Orthonormal functions	en
dc.subject.other	Physiological systems	en
dc.subject.other	Polynomial models	en
dc.subject.other	Posterior distributions	en
dc.subject.other	Statistical criterion	en
dc.subject.other	Structural parameters	en
dc.subject.other	System functions	en
dc.subject.other	Trial-and-error procedures	en
dc.subject.other	Volterra kernels	en
dc.subject.other	Identification (control systems)	en
dc.title	Bayesian model order selection for nonlinear system function expansions	en
heal.type	conferenceItem	en
heal.identifier.primary	10.1109/IEMBS.2008.4649623	en
heal.identifier.secondary	http://dx.doi.org/10.1109/IEMBS.2008.4649623	en
heal.identifier.secondary	4649623	en
heal.publicationDate	2008	en
heal.abstract	Orthonormal function expansions have been used extensively in the context of linear and nonlinear systems identification, since they result in a significant reduction in the number of required free parameters. In particular, Laguerre basis expansions of Volterra kernels have been used successfully for physiological systems identification, due to the exponential decaying characteristics of the Laguerre orthonormal basis and the inherent nonlinearities that characterize such systems. A critical aspect of the Laguerre expansion technique is the selection of the model structural parameters, i.e., polynomial model order, number of Laguerre functions in the expansion and value of the Laguerre parameter α, which determines the rate of exponential decay. This selection is typically made by trial-and-error procedures on the basis of the model prediction error. In the present paper, we formulate the Laguerre expansion technique in a Bayesian framework and derive analytically the posterior distribution of the a parameter, as well as the model evidence, in order to infer on the expansion structural parameters. We also demonstrate the performance of the proposed method by simulated examples and compare it to alternative statistical criteria for model order selection. © 2008 IEEE.	en
heal.journalName	Proceedings of the 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS'08 - ""Personalized Healthcare through Technology""	en
dc.identifier.doi	10.1109/IEMBS.2008.4649623	en
dc.identifier.volume	2008	en
dc.identifier.spage	2165	en
dc.identifier.epage	2168	en