Deviance Information Criterion (DIC)

ResearchSpace Repository

Show simple item record

dc.contributor.author Meyer, Renate en
dc.contributor.editor Balakrishnan, N en
dc.contributor.editor Brandimarte, P en
dc.contributor.editor Everitt, B en
dc.contributor.editor Molenberghs, G en
dc.contributor.editor Piegorsch, W en
dc.contributor.editor Ruggeri, F en
dc.date.accessioned 2017-04-12T02:18:51Z en
dc.date.issued 2016 en
dc.identifier.citation In Wiley StatsRef: Statistics Reference Online. Editors: Balakrishnan N, Brandimarte P, Everitt B, Molenberghs G, Piegorsch W, Ruggeri F . 5 pages. Wiley, 2016 en
dc.identifier.isbn 9781118445112 en
dc.identifier.uri http://hdl.handle.net/2292/32572 en
dc.description.abstract The deviance information criterion (DIC) was introduced in 2002 by Spiegelhalter et al. [1] to compare the relative t of a set of hierarchical Bayesian models. It is similar to Akaike's information criterion (AIC) in combining a measure of goodness-of- t and measure of complexity, both based on the deviance. While AIC uses the maximum likelhood estimate, DIC's plug-in estimate is based on the posterior mean. Since the number of independent parameters in a Bayesian hierarchical model is not clearly de ned, DIC estimates the e ective number of parameters by the di erence of the posterior mean of the deviance and the deviance at the posterior mean. This coincides with the number of independent parameters in xed e ect models with at priors, thus the DIC is a generalization of AIC. It can be justi ed as an estimate of the posterior predictive model performance within a decision-theoretic framework and it is asymptotically equivalent to leave-one-out cross- validation. The DIC has been used extensively for practical model comparison in many disciplines, works well for exponential family models but due to its dependence on the parametrization and focus of a model, its application to mixture models is problematic. en
dc.publisher Wiley en
dc.relation.ispartof Wiley StatsRef: Statistics Reference Online en
dc.rights Items in ResearchSpace are protected by copyright, with all rights reserved, unless otherwise indicated. Previously published items are made available in accordance with the copyright policy of the publisher. en
dc.rights.uri https://researchspace.auckland.ac.nz/docs/uoa-docs/rights.htm en
dc.title Deviance Information Criterion (DIC) en
dc.type Book Item en
dc.identifier.doi 10.1002/9781118445112.stat07878 en
dc.rights.holder Copyright: Wiley en
pubs.author-url http://onlinelibrary.wiley.com/book/10.1002/9781118445112.stat07878 en
dc.rights.accessrights http://purl.org/eprint/accessRights/RestrictedAccess en
pubs.elements-id 506189 en
pubs.org-id Science en
pubs.org-id Statistics en
pubs.record-created-at-source-date 2015-11-25 en
pubs.online-publication-date 2016-08-05 en


Files in this item

Find Full text

This item appears in the following Collection(s)

Show simple item record

Share

Search ResearchSpace


Advanced Search

Browse

Statistics