Sequential Likelihood Rules for Selecting the Negative Exponential Population Having the Smallest Scale Parameter

Abstract

We consider k independent two­parameter exponential populations having unknown scale parameters. The problem is to select the popula­tion associated with the smallest scale parameter under Bechhofers (1954) “indifference­zone“ approach. Likelihood based sequential rules are studied when the location parameters are unknown, known, and some known/some unknown. The general methodology has been illustrated extensively, both theoretically and by means of simulated experimunts when k=2,