search query: @indexterm STATISTICAL METHODS / total: 745
reference: 370 / 745
Author: | Koehler, G. J. |
Title: | Characterization of unacceptable solutions in LP discriminant analysis. |
Journal: | Decision Sciences
1989 : SPRING, VOL. 20:2, p. 239-257 |
Index terms: | LINEAR PROGRAMMING STATISTICAL METHODS |
Language: | eng |
Abstract: | Discriminant analysis is used to study differences between two or more mutually exclusive groups as determined by one or more variables and to predict in which group new observations are likely belong. Many linear programming models have been proposed for doing discrimination analysis. Partial characterizations for unacceptable solutions have been discussed and new models proposed to avoid these problems. The present paper characterizes those conditions which lead to unacceptable solutions in all two-group models. |
SCIMA