Computation for intrinsic variable selection in normal regression models via expected-posterior prior

Fouskakis, D; Ntzoufras, I

dc.contributor.author	Fouskakis, D	en
dc.contributor.author	Ntzoufras, I	en
dc.date.accessioned	2014-03-01T11:46:40Z
dc.date.available	2014-03-01T11:46:40Z
dc.date.issued	2012	en
dc.identifier.issn	09603174	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/38015
dc.subject	Bayesian variable selection	en
dc.subject	Expected-posterior priors	en
dc.subject	Imaginary data	en
dc.subject	Intrinsic priors	en
dc.subject	Jeffreys prior	en
dc.subject	Normal regression models	en
dc.subject	Objective model selection methods	en
dc.title	Computation for intrinsic variable selection in normal regression models via expected-posterior prior	en
heal.type	other	en
heal.identifier.primary	10.1007/s11222-012-9325-9	en
heal.identifier.secondary	http://dx.doi.org/10.1007/s11222-012-9325-9	en
heal.publicationDate	2012	en
heal.abstract	In this paper, we focus on the variable selection problem in normal regression models using the expected-posterior prior methodology. We provide a straightforward MCMC scheme for the derivation of the posterior distribution, as well as Monte Carlo estimates for the computation of the marginal likelihood and posterior model probabilities. Additionally, for large spaces, a model search algorithm based on {Mathematical expression} is constructed. The proposed methodology is applied in two real life examples, already used in the relevant literature of objective variable selection. In both examples, uncertainty over different training samples is taken into consideration. © 2012 Springer Science+Business Media, LLC.	en
heal.journalName	Statistics and Computing	en
dc.identifier.doi	10.1007/s11222-012-9325-9	en
dc.identifier.spage	1	en
dc.identifier.epage	9	en