Please use this identifier to cite or link to this item:
Title: Bayesian Monte Carlo testing with one-dimensional measures of evidence.
Authors: Silva, Ivair Ramos
Marques, Reinaldo Antônio Gomes
Keywords: Bayes factor
Expected loss
Exact test
Issue Date: 2019
Citation: SILVA, I. R.; MARQUES, R. A. G. Bayesian Monte Carlo testing with one-dimensional measures of evidence. Journal of Computational and Applied Mathematics, v. 351, p. 250-259, maio 2018. Disponível em: <>. Acesso em: 19 mar. 2019.
Abstract: Bayesian hypothesis testing procedures are constructed by means of test statistics which are functions of the posterior distribution. Usually, the whole sample vector is selected to form the sufficient empirical part of the posterior distribution. But, in certain problems, one may prefer to use well-established one-dimensional sufficient statistics in place of the sample vector. This paper introduces a Bayesian Monte Carlo procedure specially designed for such cases. It is shown that the performance of this new approach is arbitrarily close to the exact Bayesian test. In addition, for arbitrary desired precisions, we develop a theoretical rule of thumb for choosing the minimum number m0 of Monte Carlo simulations. Surprisingly, m0 does not depend on the shape of loss/cost functions when those are used to compound the test statistic. The method is illustrated for testing mean vectors in highdimension and for detecting spatial clusters of diseases in aggregated maps.
ISSN: 0377-0427
Appears in Collections:DEEST - Artigos publicados em periódicos

Files in This Item:
File Description SizeFormat 
  Restricted Access
570,87 kBAdobe PDFView/Open

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