Enhancing class discrimination in Kernel Discriminant Analysis

Alexandros Iosifidis, Anastasios Tefas, Ioannis Pitas

Research output: Chapter in Book/Report/Conference proceedingConference Contribution (Conference Proceeding)

221 Downloads (Pure)

Abstract

In this paper, we propose an optimization scheme aiming at optimal nonlinear data projection, in terms of Fisher ratio maximization. To this end, we formulate an iterative optimization scheme consisting of two processing steps: optimal
data projection calculation and optimal class representation determination. Compared to the standard approach employing the class mean vectors for class representation, the proposed optimization scheme increases class discrimination in the reduced-dimensionality feature space. We evaluate the proposed method in standard classification problems, as well as on the classification of human actions and face, and show that it is able to achieve better generalization performance, when compared to the standard approach.
Original languageEnglish
Title of host publication2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2015)
Subtitle of host publicationProceedings of a meeting held 19-24 April 2015, South Brisbane, Queensland, Australia
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Pages1926-1930
Number of pages5
ISBN (Electronic)9781467369978
ISBN (Print)9781467369985
DOIs
Publication statusPublished - Sep 2015
Event2015 (40th) IEEE International Conference on Acoustics, Speech and Signal Processing - Brisbane, Australia
Duration: 19 Apr 201524 Apr 2015

Publication series

NameProceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
ISSN (Print)1520-6149

Conference

Conference2015 (40th) IEEE International Conference on Acoustics, Speech and Signal Processing
Abbreviated titleICASSP '15
CountryAustralia
CityBrisbane
Period19/04/1524/04/15

Keywords

  • Kernel Discriminant Analysis
  • Optimized Class Representation
  • Nonlinear data projection

Fingerprint Dive into the research topics of 'Enhancing class discrimination in Kernel Discriminant Analysis'. Together they form a unique fingerprint.

Cite this