On the choice of variables in classification problems with dichotomous variables

1 December 1967

journal article
Published by Oxford University Press (OUP) in Biometrika

Vol. 54 (3-4) , 668-670
https://doi.org/10.1093/biomet/54.3-4.668

Abstract

We study how to choose two out of p dichotomous variables in classification problems. Our results show that positive correlation may increase discrimination while negative correlation may decrease discrimination. These results were not predicted from the analysis of the same problems with normally distributed variables.

Keywords

DICHOTOMOUS VARIABLES
CLASSIFICATION PROBLEMS
VARIABLES IN CLASSIFICATION
PROBLEMS WITH DICHOTOMOUS
CHOICE OF VARIABLES

This publication has 0 references indexed in Scilit: