D. Salmerón Martínez, J. A. Cano Sánchez, C. P. Robert
Integral priors were developed for Bayesian model selection and have been successfully applied in many situations. However, there are two aspects that deserve special attention. First, the method is stated for the comparison of two models. Second, nonparametric density estimates of the integral priors have been typically needed to approximate the Bayes factors, which translates into more computing time. Here we generalize the definition for more than two models and propose new numerical procedures to approximate the Bayes factors. The method is illustrated with several examples including location-scale models, Poisson versus the negative binomial family, hypothesis testing for the exponential distribution mean, and the problem of testing if the mean of the normal distribution with unknown variance is negative, zero, or positive. Finally we illustrate the method for the variable selection problem.
Keywords: Integral priors, Bayesian model selection, Objective Bayes factor, Markov chains.
Scheduled
GT21 Bayesian Inference II
June 7, 2022 3:30 PM
Audiovisual room