On prior-data conflict in predictive Bernoulli inferences

G.M. Walter, Th. Augustin, F.P.A. Coolen

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

6 Citations (Scopus)
24 Downloads (Pure)

Abstract

By its capability to deal with the multidimensional nature of uncertainty, imprecise probability provides a powerful methodology to sensibly handle prior-data conflict in Bayesian inference. When there is strong conflict between sample observations and prior knowledge the resulting posterior model naturally should be much more imprecise than in the situation of mutual agreement or compatibility. Focusing presentation on the prototypical example of Bernoulli trials, we discuss the ability of different approaches to deal with prior-data conflict. We study a generalized Bayesian setting, including Walley's Imprecise Beta-Binomial model and his extension to handle prior data conflict (called pdc-IBBM here). We investigate alternative shapes of prior parameter sets, chosen in a way that shows improved behaviour in the case of prior-data conflict and their influence on the posterior predictive distribution. Thereafter we present a new approach, consisting of an imprecise weighting of two originally separate inferences, one of which is based on an informative imprecise prior whereas the other one is based on an uninformative imprecise prior. This approach deals with prior-data conflict in a fascinating way.
Original languageEnglish
Title of host publicationISIPTA'11
Subtitle of host publicationProceedings of the Seventh International Symposium on Imprecise Probabilities: Theories and Applications
EditorsFrank P.A. Coolen, Gert de Cooman, Thomas Fetz, Michael Oberguggenberger
Place of PublicationInnsbruck
PublisherSociety for Imprecise Probability: Theories and Applications
Pages391-400
Number of pages10
Publication statusPublished - Jul 2011
Externally publishedYes
EventISIPTEA'11, Seventh International Symposium on Imprecise Probabilities: Theories and Applications - Innsbruck, Austria
Duration: 25 Jul 201128 Jul 2011

Conference

ConferenceISIPTEA'11, Seventh International Symposium on Imprecise Probabilities: Theories and Applications
Country/TerritoryAustria
CityInnsbruck
Period25/07/1128/07/11

Keywords

  • Bayesian inference
  • generalized iLUCK-models
  • imprecise Beta-Binomial model
  • imprecise weighting
  • predictive inference
  • prior-data conflict

Fingerprint

Dive into the research topics of 'On prior-data conflict in predictive Bernoulli inferences'. Together they form a unique fingerprint.

Cite this