E. Pleszczynska, M. Niewiadomska-Bugaj, T. Kowalczyk

Grade counterpart of classical data analysis

882

Abstract

Grade statistical methods are applicable to analysis of data on any measurment scale. These techniques are especially usefull in solving problems related to the dependence structure in bivariate (multivariate) data. We begin with introduction and discussion of the main concept, that of a concentration index in pairs of probability distributions. Later we present various methods - among them grade cluster analysis and grade correspondence analysis - with applications to practical problems and real data sets. Finally, we provide a comprehensive list of related publications.

Key words: concentration index, clustering, decomposition, Kendall's tau, Sperman's rho.