Posterior analysis of lognormal regression models using the Gibbs sampler

摘要

The Gibbs sampler technique has been found to be very useful for extracting samples from the complex posterior forms and the procedure is equally well applicable in teh analysis of both the complete and the censored data. In this paper we consider the lognormal model, an important model considered in wide variety of problems, and assume that the regressor (or stress) variables affect the two parameters of the distribution in a number of ways.