The Bayesian quantile regression method in accelerated life tests

ABSTRACT In accelerated life tests (ALT), there are many limitations for a traditional parametric method to estimate a particular quantile in the lower tail of the lifetime distribution at the use condition. This article develops a Bayesian quantile regression method for analyzing ALT data. As a distribution-free method, Bayesian quantile regression method is robust to lifetime distribution uncertainty and flexible to deal with the relationship between shape parameter and accelerating stresses. A Gibbs sampling method is implemented for Bayesian posterior inference. This method is easy to implement and allows us to use data augmentation to simplify the computation in the presence of right-censored observations. A simulation study and a real example are used to demonstrate the advanced properties of proposed method.

[1]  Michael S. Hamada,et al.  Bayesian Analysis of Step-Stress Accelerated Life Tests and Its Use in Planning , 2015 .

[2]  Shuguang He,et al.  Reliability Modeling of Accelerated Life Tests with Both Random Effects and Nonconstant Shape Parameters , 2015 .

[3]  I-Tang Yu,et al.  Applying Bayesian Model Averaging for Quantile Estimation in Accelerated Life Tests , 2012, IEEE Transactions on Reliability.

[4]  W. Meeker.  Accelerated Testing: Statistical Models, Test Plans, and Data Analyses , 1991 .

[5]  Wayne Nelson,et al.  Analysis of Accelerated Life Test Data-Least Squares Methods for the Inverse Power Law Model , 1975, IEEE Transactions on Reliability.

[6]  Ancha Xu,et al.  Reference optimality criterion for planning accelerated life testing , 2015 .

[7]  O. I. Tyoskin,et al.  Nonparametric model for step-stress accelerated life testing , 1996, IEEE Trans. Reliab..

[8]  E. Elsayed,et al.  A general accelerated life model for step-stress testing , 2005 .

[9]  Cristina Davino,et al.  Quantile Regression: Theory and Applications , 2013 .

[10]  William Q. Meeker,et al.  Optimum Accelerated Life Tests Wth a Nonconstant Scale Parameter , 1994 .

[11]  Nan Chen,et al.  Robust Quantile Analysis for Accelerated Life Test Data , 2016, IEEE Transactions on Reliability.

[12]  M. Hamada,et al.  Analysis of Censored Data from Fractionated Experiments: A Bayesian Approach , 1995 .

[13]  T. Lancaster,et al.  Bayesian quantile regression methods , 2010 .

[14]  Yada Zhu,et al.  Optimal design of accelerated life testing plans under progressive censoring , 2013 .

[15]  Freda Kemp,et al.  The Laplace Distribution and Generalizations: a Revisit with Applications to Communications, Economics, Engineering, and Finance , 2003 .

[16]  D. Bates,et al.  Output Analysis and Diagnostics for MCMC , 2015 .

[17]  Keming Yu,et al.  Bayesian quantile regression , 2001 .

[18]  S. Walker.  Invited comment on the paper "Slice Sampling" by Radford Neal , 2003 .

[19]  Richard L. Schmoyer,et al.  Nonparametric analyses for two-level single-stress accelerated life tests , 1991 .

[20]  E. Tsionas.  Bayesian quantile inference , 2003 .

[21]  Francis G. Pascual,et al.  Model-Robust Test Plans With Applications in Accelerated Life Testing , 2003, Technometrics.

[22]  P. W. Srivastava,et al.  Optimum step-stress partially accelerated life tests for the truncated logistic distribution with censoring , 2010 .

[23]  Francis G. Pascual,et al.  Accelerated Life Test Plans Robust to Misspecification of the Stress—Life Relationship , 2006, Technometrics.

[24]  Keming Yu,et al.  A partially collapsed Gibbs sampler for Bayesian quantile regression , 2009 .

[25]  H. Kozumi,et al.  Gibbs sampling methods for Bayesian quantile regression , 2011 .

[26]  N. Shephard,et al.  Non‐Gaussian Ornstein–Uhlenbeck‐based models and some of their uses in financial economics , 2001 .