A COMPARISON OF ERROR PROBABILITIES FOR VARIOUS DISCRIMINATION PROCEDURES WHEN THE POPULATIONS ARE GAMMA.
Abstract
The linear discriminant function is a widely employed classification procedure, therefore, it is of interest to determine the performance of this procedure when the populations are not gaussian. A complete comparison is provided of the probabilities of misclassification of a nonparametric procedure against the linear discriminant function when the two populations are exponential. A limited tabulation is included.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1964
- Accession Number
- AD0481205
Entities
People
- Charles F. Hager
Organizations
- Naval Postgraduate School