Premium
A non‐parametric classification rule for several multivariate populations
Author(s) -
Kanazawa Mitsuyo
Publication year - 1974
Publication title -
canadian journal of statistics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.804
H-Index - 51
eISSN - 1708-945X
pISSN - 0319-5724
DOI - 10.2307/3314689
Subject(s) - statistic , multivariate statistics , consistency (knowledge bases) , mathematics , parametric statistics , statistics , classification rule , wilcoxon signed rank test , population , pattern recognition (psychology) , artificial intelligence , computer science , discrete mathematics , medicine , mann–whitney u test , environmental health
Abstract In a classification problem such that a set of observations is classified into a population when it is known beforehand that the observations have come from one of k (>2) multivariate populations, a non‐parametric classification rule based on the Wilcoxon type statistic is proposed, and its consistency is shown. It is also shown that the statistic is asymptotically distributed according to the chi‐square distribution with p degrees of freedom when the observations are classified correctly.