haku: @author Koehler, G. J. / yhteensä: 5
viite: 2 / 5
Tekijä: | Koehler, G. J. Erenguc, S. S. |
Otsikko: | Minimizing misclassifications in linear discriminant analysis |
Lehti: | Decision Sciences
1990 : WINT. VOL. 21:1, p.63-83 |
Asiasana: | MATHEMATICAL PROGRAMMING MIXED INTEGER PROGRAMMING |
Kieli: | eng |
Tiivistelmä: | Discriminant analysis is used for two purposes: (1) to classify observations into mutually exclusive groups and (2) to explain differences between these groups based on observable attributes. The present paper develops a procedure for determining two-group linear discriminant classifiers that misclassify the minimal number of observations in the training examples. An experimental study confirms the value of the approach. |
SCIMA