Linear discriminant analysis for two classes via recursive neural network reduction of the class separation

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

A method for the linear discrimination of two classes is presented. It maximizes the Patrick-Fisher (PF) distance between the projected class-conditional densities. Since the PF distance is a highly nonlinear function, we propose a method, which searches for the directions corresponding to several large local maxima of the PF distance. Its novelty lies in a neural network transformation of the data along a found direction into data with deflated maxima of the PF distance and iteration to obtain the next direction. A simulation study indicates that the method has the potential to find the global maximum of the PF distance.

Original languageEnglish
Title of host publicationAdvances in Pattern Recognition - Joint IAPR International Workshops SSPR 1998 and SPR 1998, Proceedings
EditorsAdnan Amin, Dov Dori, Pavel Pudil, Herbert Freeman
PublisherSpringer Verlag
Pages775-784
Number of pages10
ISBN (Print)3540648585, 9783540648581
DOIs
StatePublished - 1 Jan 1998
Event7th Joint IAPR International Workshop on Structural and Syntactic Pattern Recognition, SSPR 1998 and 2nd International Workshop on Statistical Techniques in Pattern Recognition, SPR 1998 - Sydney, Australia
Duration: 11 Aug 199813 Aug 1998

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1451
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference7th Joint IAPR International Workshop on Structural and Syntactic Pattern Recognition, SSPR 1998 and 2nd International Workshop on Statistical Techniques in Pattern Recognition, SPR 1998
Country/TerritoryAustralia
CitySydney
Period11/08/9813/08/98

Keywords

  • Auto-associative network
  • Discriminant analysis
  • Neural networks for classification
  • Projection pursuit
  • Statistical pattern recognition

Fingerprint

Dive into the research topics of 'Linear discriminant analysis for two classes via recursive neural network reduction of the class separation'. Together they form a unique fingerprint.

Cite this