A Counter-Example in Linear Feature Selection Theory

Abstract
A family of examples is constructed to show that there may exist distinct k × n matrices, each of which maximizes transformed divergence which are not row-equivalent.

This publication has 1 reference indexed in Scilit: