Bayesian Semiparametric Symmetric Models for Binary Data

Marcio Augusto Diniz, Carlos Alberto de Braganca Pereira, Adriano Polpo

Research output: Chapter in Book/Conference paperConference paper

Abstract

This work proposes a general Bayesian semiparametric model for binary data. Symmetric prior probability curves as an extension for discussed ideas from Basu and Mukhopadhyay (Generalized Linear Models: A Bayesian Perspective, pp. 231-241, 1998) are considered using the blocked Gibbs sampler, which is more general than the Polya urn Gibbs sampler. The Bayesian semiparametric approach allows us to incorporate uncertainty around the F distribution of the latent data and to model heavy-tailed or light-tailed distributions. In particular, the Bayesian semiparametric logistic model is introduced, which enables one to elicit prior distributions for regression coefficients from information about odds ratios; this is quite interesting in applied research. Then, this framework opens several possibilities to deal with binary data in the Bayesian perspective.

Original languageEnglish
Title of host publicationINTERDISCIPLINARY BAYESIAN STATISTICS, EBEB 2014
EditorsA Polpo, F Louzada, LLR Rifo, JM Stern, M Lauretto
PublisherSpringer
Pages323-335
Number of pages13
ISBN (Print)978-3-319-12453-7
DOIs
Publication statusPublished - 2015
Externally publishedYes
Event12th Brazilian Meeting on Bayesian Statistics (EBEB) - Hotel Fazenda Hípica Atibaia, Atibaia, Brazil
Duration: 10 Mar 201414 Mar 2014
https://www.ime.usp.br/~isbra/ebeb/ebeb2014/

Publication series

NameSpringer Proceedings in Mathematics & Statistics
PublisherSPRINGER
Volume118
ISSN (Print)2194-1009

Conference

Conference12th Brazilian Meeting on Bayesian Statistics (EBEB)
CountryBrazil
CityAtibaia
Period10/03/1414/03/14
Otherhe Brazilian Meeting on Bayesian Statistics (EBEB) is in its twelfth edition. This series of meetings aims at strengthening the research on Bayesian methods and widening their application. It also provides an environment where Brazilian and international researchers collaborate, present their most recent developments and discuss on open problems. EBEB also allows graduate students to make contacts with experienced researchers. This year's meeting has a particular focus on discussing recent developments in the many viewpoints of Bayesian statistics, such as computational, theoretical, methodological and applied views.
Internet address

Cite this

Diniz, M. A., de Braganca Pereira, C. A., & Polpo, A. (2015). Bayesian Semiparametric Symmetric Models for Binary Data. In A. Polpo, F. Louzada, LLR. Rifo, JM. Stern, & M. Lauretto (Eds.), INTERDISCIPLINARY BAYESIAN STATISTICS, EBEB 2014 (pp. 323-335). (Springer Proceedings in Mathematics & Statistics; Vol. 118). Springer. https://doi.org/10.1007/978-3-319-12454-4_27