Bayesian Type I Censored Designs for the Weibull Lifetime Model: Information Based Criterion
Zaher, A., Ismail, M., Bahaa, M. (1996). Bayesian Type I Censored Designs for the Weibull Lifetime Model: Information Based Criterion. EKB Journal Management System, 40(2), 127-150. doi: 10.21608/esju.1996.314786
Adel M. Zaher; Moshira A. Ismail; Mohamed S. Bahaa. "Bayesian Type I Censored Designs for the Weibull Lifetime Model: Information Based Criterion". EKB Journal Management System, 40, 2, 1996, 127-150. doi: 10.21608/esju.1996.314786
Zaher, A., Ismail, M., Bahaa, M. (1996). 'Bayesian Type I Censored Designs for the Weibull Lifetime Model: Information Based Criterion', EKB Journal Management System, 40(2), pp. 127-150. doi: 10.21608/esju.1996.314786
Zaher, A., Ismail, M., Bahaa, M. Bayesian Type I Censored Designs for the Weibull Lifetime Model: Information Based Criterion. EKB Journal Management System, 1996; 40(2): 127-150. doi: 10.21608/esju.1996.314786

Bayesian Type I Censored Designs for the Weibull Lifetime Model: Information Based Criterion

The Egyptian Statistical Journal
Article 2, Volume 40, Issue 2, December 1996, Page 127-150  XML PDF (34.21 MB)
Document Type: Original Article
DOI: 10.21608/esju.1996.314786
View on SCiNiTO View on SCiNiTO
Authors
Adel M. Zaher* 1; Moshira A. Ismail2; Mohamed S. Bahaa2
1Faculty of Economics & Political Science, Cairo University, Egypt
2Faculty of Economics & Political Science, Cairo University, Egypt.
Abstract
In this paper, Bayesian life test sampling plans are developed for the Weibull distribution with known shape parameter under type I censoring. Specifically, we consider the choice of the truncation time to and the sample size n which maximize the expected gain in information under a certain cost constraint. This measure of information is appropriate when the objective is to gain knowledge about a certain parameter or a future observation. We mainly restrict our attention to deal with information about the scale parameter of the distribution and any-one-to-one transformation of it. The exact evaluation of the expression of information and the corresponding optimal design is obtained numerically. Approximate expressions for the expected gain in information are suggested and used to derive some analytical results for the optimal design. Numerical results are provided to check the accuracy of the approximate results. A related problem where information is required about the minimum lifetime of a future experiment is also discussed.
Keywords
Bayesian Analysis; Type I Censoring; Weibull Distribution; Information; Optimal Design; Cost Constraint
Statistics
Article View: 95
PDF Download: 17