Discriminatory Analysis - Nonparametric Discrimination: Small Sample Performance

Abstract

A classification procedure is worked out for the following situations: Two large samples, one from each of two populations, have been observed. An individual of unknown origin is to be classified as belonging to the first population if the majority of a specified odd number of individuals closet to the individual in question belong to the first population. This method has optimum properties when the number of closest individuals is permitted to be very large. For certain cases involving multivariate normal distributions with the same covariance matrix, the probabilities of possible misclassification have been computed and compared with those of the discriminant function method.