Journal:Informatica
Volume 3, Issue 3 (1992), pp. 348–359
Abstract
The rules of classification of the group of N independent observations into one of k normal populations are considered. In the case when parameters are not known, Bayes estimators of the discriminant scores are suggested.