ABSTRACT

In this chapter, the authors consider two distinct exponential populations and discuss the classification of a new observation, which is known to have come from one of these two populations, as belonging to one of the two populations. The theory of classification is of considerable interest to practitioners, and it is used to categorize an object (or an individual) on the basis of a profile of its characteristics. Most of the statistical developments in classification theory have primarily been on the normal (univariate or multivariate) populations. However, in many life-testing experiments one may often observe a Type-II right censored sample. The authors derive the likelihood-ratio classification procedure and then propose a plug-in procedure and presents exact expressions for the errors of the plug-in procedure.