Elsevier

Pattern Recognition Letters

Volume 15, Issue 11, November 1994, Pages 1119-1125
Pattern Recognition Letters

Floating search methods in feature selection

https://doi.org/10.1016/0167-8655(94)90127-9Get rights and content

Abstract

Sequential search methods characterized by a dynamically changing number of features included or eliminated at each step, henceforth “floating” methods, are presented. They are shown to give very good results and to be computationally more effective than the branch and bound method.

References (16)

  • P. Pudil et al.

    An analysis of the max-min approach to feature selection

    Pattern Recognition Lett.

    (1993)
  • W. Siedlecki et al.

    A note on genetic algorithm for large-scale feature selection

    Pattern Recognition Lett.

    (1989)
  • E. Backer et al.

    On the max-min approach for feature ordering and selection

  • N. Choakjarernwanit

    Feature selection in pattern recognition

    Technical Report VSSP-TR-1/91, University of Surrey, UK

    (1991)
  • N. Choakjarernwanit et al.

    A detailed study of the max-min in pattern recognition

    Technical Report VSSP-TR-5/91, University of Surrey, UK

    (1991)
  • M. Ichino et al.

    Optimum feature selection by zero-one integer programming

    IEEE Trans. Syst. Man Cybernet.

    (1984)
  • J. Kittler

    Feature set search algorithms

  • T. Marill et al.

    On the effectiveness of receptors in recognition system

    IEEE Trans. Inform. Theory

    (1963)
There are more references available in the full text version of this article.

Cited by (2589)

View all citing articles on Scopus
1

Permanently with the Institute of Information Theory and Automation, Academy of Sciences of the Czech Republic, 182 08 Prague 8, Czech Republic.

View full text