Portal de Periódicos da CAPES

Artigo Produção Nacional Revisado por pares

BIBTEX RIS

Mixture of Normal Distributions in Multivariate Null Intercept Measurement Error Model

2006; Taylor & Francis; Volume: 16; Issue: 6 Linguagem: Inglês

10.1080/10543400600801265

1520-5711

Reiko Aoki, Dorival Leão Pinto Júnior, Jorge Alberto Achcar, Heleno Bolfarine,

Bayesian Methods and Mixture Models

In this paper we propose the use of a multivariate null intercept measurement error model, where the true unobserved value of the covariate follows a mixture of two normal distributions. The proposed model is applied to a dental clinical trial presented in Hadgu and Koch (1999). A Bayesian approach is considered and a Gibbs Sampler is used to perform the computations.