Adventures in Discriminant Analysis

Jeffrey L. Solka
Naval Surface Warfare Center, Dahlgren Division
and
George Mason University, Institute for Computational Science and Informatics


ABSTRACT

This talk will describe our recent work in a multi-class discriminant analysis problem which utilizes relatively high dimensional features. Topics to be discussed will include exploratory data analysis, optimal feature selection, density estimation, and classifier performance measures. Difficulties encountered during the analysis will be discussed.