Translator Disclaimer
September 2021 A Symmetric Prior for Multinomial Probit Models
Lane F. Burgette, David Puelz, P. Richard Hahn
Author Affiliations +
Bayesian Anal. 16(3): 991-1008 (September 2021). DOI: 10.1214/20-BA1233

Abstract

Fitted probabilities from widely used Bayesian multinomial probit models can depend strongly on the choice of a base category, which is used to uniquely identify the parameters of the model. This paper proposes a novel identification strategy, and associated prior distribution for the model parameters, that renders the prior symmetric with respect to relabeling the outcome categories. The new prior permits an efficient Gibbs algorithm that samples rank-deficient covariance matrices without resorting to Metropolis-Hastings updates.

Citation

Download Citation

Lane F. Burgette. David Puelz. P. Richard Hahn. "A Symmetric Prior for Multinomial Probit Models." Bayesian Anal. 16 (3) 991 - 1008, September 2021. https://doi.org/10.1214/20-BA1233

Information

Published: September 2021
First available in Project Euclid: 18 August 2020

Digital Object Identifier: 10.1214/20-BA1233

JOURNAL ARTICLE
18 PAGES


SHARE
Vol.16 • No. 3 • September 2021
Back to Top