Sufficient conditions for optimality of a \((z,c^ -,c^ +)\)-sampling plan in multistage Bayesian acceptance sampling (Q1091730)

This paper provides a general framework for designing a multistage acceptance sampling plan so as to minimize the Bayesian expected total cost of handling the noninspected part of the lot. It is proved that, under certain intuitive conditions, the sampling plan \((z,c^-,c^+)\) is optimal. The structure of this plan is also discussed.