Abstract
Generalized hybrid censoring schemes, proposed by Chandrasekar et al. (Naval Research Logistics, 51(7), 994–1004, 2004), have several advantages over the conventional hybrid censoring schemes. In this paper, we introduce a new generalized Type-II hybrid censoring scheme for two samples. The maximum likelihood and Bayesian inferential approaches for estimating the unknown mean lifetimes of the experimental units for the two samples follow exponential population with different scale parameters are considered. The corresponding asymptotic confidence intervals of the maximum likelihood estimators are also obtained. Using gamma conjugate priors, the Bayes estimators are developed relative to both symmetric and asymmetric loss functions. Also, some popular censoring plans are generalized and can be obtained as a special cases from our results. One real-life data set is analyzed to discuss how the applicability of the proposed methods in real phenomenon. Finally, to examine the performance of proposed methods, a Monte Carlo simulation study is carried out.