|
|||||
|
|
An extension of geisser's discrimination model to proportional covariance matrices |
|
| Authors: | Douglas M Hawkins E Liefde Raath |
| Abstract: | The predictive discrimination model of Geisser is derived for the case in which the covariance matrices of the different populations are assumed proportional, with unknown constants of proportionality. It is shown that this model provides a better fit to a set of data than does the usual homoscedastic model. |
| Keywords: | Bayesian predictive discrimination discriminant analysis heteroscedasticity |