Cognitive Computation

, Volume 2, Issue 3, pp 160–164

Reducing Features Using Discriminative Common Vectors

  • Carlos M. Travieso
  • Marcos del Pozo
  • Miguel A. Ferrer
  • Jesús B. Alonso
Article

DOI: 10.1007/s12559-010-9059-y

Cite this article as:
Travieso, C.M., del Pozo, M., Ferrer, M.A. et al. Cogn Comput (2010) 2: 160. doi:10.1007/s12559-010-9059-y

Abstract

A feature reduction system based on Discriminative Common Vector is presented and evaluated in this paper. The validation of this system was made with three databases, first one is DNA markers and the other two are The ORL Database of Face and The Yale Face Database. Moreover, a supervised classification system has been implemented with three different classifiers, achieving the best success rates with Support Vector Machines using Radial Basis Function kernel and a one-versus-all multiclass approach. The study shows clearly how our approach reduces the number of features and load times, keeping or improving the level of discrimination.

Keywords

Reduction featuresDiscriminative Common VectorMachine learningPattern Recognition

Copyright information

© Springer Science+Business Media, LLC 2010

Authors and Affiliations

  • Carlos M. Travieso
    • 1
  • Marcos del Pozo
    • 1
  • Miguel A. Ferrer
    • 1
  • Jesús B. Alonso
    • 1
  1. 1.Signals and Communications Department, IDeTICUniversity of Las Palmas de Gran CanariaLas Palmas de Gran CanáriaSpain