Please use this identifier to cite or link to this item:
Full metadata record
DC FieldValueLanguage
dc.contributor.authorCorcuera Valverde, José Manuelcat
dc.contributor.authorGiummolè, Federicacat
dc.description.abstractIn a recent paper, Komaki studied the second-order asymptotic properties of predictive distributions, using the Kullback-Leibler divergence as a loss function. He showed that estimative distributions with asymptotically efficient estimators can be improved by predictive distributions that do not belong to the model. The model is assumed to be a multidimensional curved exponential family. In this paper we generalize the result assuming as a loss function any f divergence. A relationship arises between alpha connections and optimal predictive distributions. In particular, using an alpha divergence to measure the goodness of a predictive distribution, the optimal shift of the estimate distribution is related to alpha-covariant derivatives. The expression that we obtain for the asymptotic risk is also useful to study the higher-order asymptotic properties of an estimator, in the mentioned class of loss functions.eng
dc.format.extent14 p.-
dc.publisherBernoulli Society for Mathematical Statistics and Probability-
dc.relation.isformatofReproducció del document publicat a:
dc.relation.ispartofBernoulli, 1999, vol. 5, núm. 1, p. 163-176-
dc.rights(c) ISI/BS, International Statistical Institute, Bernoulli Society, 1999-
dc.subject.classificationGeometria diferencialcat
dc.subject.classificationConnexions (Matemàtica)cat
dc.subject.classificationEstadística matemàticacat
dc.subject.classificationTeoria de la predicciócat
dc.subject.otherDifferential geometryeng
dc.subject.otherPrediction theoryeng
dc.subject.otherConnections (Mathematics)eng
dc.subject.otherMathematical statisticscat
dc.titleOn the relationship between connections and the asymptotic properties of predictive distributionseng
Appears in Collections:Articles publicats en revistes (Matemàtiques i Informàtica)

Files in This Item:
File Description SizeFormat 
141404.pdf303.49 kBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.