Alternative Local Discriminant Bases Using Empirical Expectation and Variance Estimation

TL;DR

This paper introduces new discriminant measures for basis selection in classification, generalizing existing algorithms and demonstrating improved performance through experiments.

Contribution

It proposes alternative discriminant measures and a generalized Local Discriminant Basis Algorithm for enhanced basis selection in classification tasks.

Findings

New discriminant measures outperform previous methods

Generalized algorithm improves basis selection accuracy

Experimental results validate the effectiveness of the proposed methods

Abstract

We propose alternative discriminant measures for selecting the best basis among a large collection of orthonormal bases for classification purposes. A generalization of the Local Discriminant Basis Algorithm of Saito and Coifman is constructed. The success of these new methods is evaluated and compared to earlier methods in experiments.