Inference for two-parameter Birnbaum-Saunders distribution based on multi-sample progressively Type-II censored data

As a generalization of the progressive Type-II censoring model, in this paper we consider the multi-sample progressive Type-II censoring model. In this work, we model the lifetimes on test with the two-parameter Birnbaum-Saunders distribution and carry out estimation of the unknown lifetime distribution parameters as well as prediction of the lifetimes of the removed units. Inference is carried out using maximum likelihood estimation and Bayesian estimation. For prediction of lifetimes, Bayesian methods are used to derive point predictors and credible intervals. Markov Chain Monte Carlo (MCMC) sampling is employed to compare the methods and one real data set is finally analyzed for illustrative purposes.