Semiparametric Model and Bayesian Analysis for ...

One of the difficulties in analyzing accelerated life testing data is the model-based failure probability prediction. Choosing an inferior model yields inaccurate predictions that can be exaggerated by extrapolation. Furthermore, testing data are often naturally clustered in groups, thus some modeling exibility must be granted to handle both the intra-cluster and inter-cluster variations. To address these problems, we discuss a data fitting strategy in this paper by developing a semiparametric model with random effects and the Bayesian piecewise exponential inference method. The proportional hazard model and Weibull accelerated failure time model are examined and compared. Our result suggests that the Bayesian piecewise exponential model with random effects outperforms other models.