Parametric and Nonparametric Discrimination for Categorical Variables
| Authors | |
|---|---|
| Year of publication | 2004 |
| Type | Article in Proceedings |
| Conference | Summer School DATASTAT 03 Folia Facultatis Scientiarum Naturalium Universitatis Masarykianae Brunensis |
| MU Faculty or unit | |
| Citation | |
| Field | Applied statistics, operation research |
| Keywords | multinomial classification; parametric and nonparametric discriminant analysis; kernel density estimation; product kernels; bandwidth choice; multivariate binary data; unordered categorical data |
| Description | The problem discussed here is that of discriminating between groups on the basis of m categorical variables. Two main approaches to discriminant analysis are explained with respect to the use of either parametric (based on multinomial distribution) or nonparametric (based on the kernel density estimation) models. An example is given to illustrate the methods. |
| Related projects: |