Optimal curtailed Bayesian sampling plans for exponential distributions with Type-I hybrid censored samples

Abstract

This paper studies a method for constructing a curtailed Bayesian sampling plan with Type-I hybrid censored samples. Based on the Bayesian sampling plan (BSP) of Liang and Yang (2013), a curtailed Bayesian sampling plan (CBSP) is constructed through an online curtailed censoring process. We show that the resulting associated curtailed Bayes decision function is equivalent to the Bayes decision function of Liang and Yang (2013). It is shown that the risk of CBSP is always less than or equal to that of BSP. Monte Carlo simulations are carried out to illustrate the performance between the two sampling plans BSP and CBSP. Numerical results illustrate that the risk of CBSP has better performance than that of BSP, and hence the CBSP is recommended.

Keywords:

• Bayesian sampling plan
• Curtailed Bayesian sampling plan
• Decision function
• Exponential distribution
• Hybrid censoring

Mathematics Subject Classification:

• 62F15
• 62C10

Acknowledgements

The authors would like to thank an Associate Editor, and a reviewer for their careful reading of the manuscript and insightful suggestions that greatly improve the presentation of the article. The work of Lee-Shen Chen was partially supported by the grant (MOST 104-2118-M-130-003-) of Ministry of Science and Technology in Taiwan.