The problem of a single sampling plan with polynomial loss for the exponential distribution based on uniformly distributed random censored data has been considered. A Bayes sampling plan is derived under various schemes of censoring time. It is specially focused on a quadratic loss and an unit time cost is included in the loss. Some optimal Bayes solutions are tabulated and some numerical comparisons between the proposed plan and a known plan under special loss are also made. It is shown that the optimal solutions of the known plan are not Bayes in general.
關聯:
Computational Statistics and Data Analysis 44(4), pp.669-691