European Symposium on Principles of Data Mining and Knowledge Discovery

PKDD 1998: Principles of Data Mining and Knowledge Discovery pp 212-220

Detection of interdependences in attribute selection

  • Javier Lorenzo
  • Mario Hernández
  • Juan Méndez
Communications Session 8. Attribute Selection

DOI: 10.1007/BFb0094822

Volume 1510 of the book series Lecture Notes in Computer Science (LNCS)
Cite this paper as:
Lorenzo J., Hernández M., Méndez J. (1998) Detection of interdependences in attribute selection. In: Żytkow J.M., Quafafou M. (eds) Principles of Data Mining and Knowledge Discovery. PKDD 1998. Lecture Notes in Computer Science, vol 1510. Springer, Berlin, Heidelberg

Abstract

A new measure for attribute selection, called GD, is proposed. The GD measure is based on Information Theory and allows to detect the interdependence between attributes. This measure is based on a quadratic form of the Mántaras distance and a matrix called Transinformation Matrix. In order to test the quality of the proposed measure, it is compared with other two feature selection methods, namely Mántaras distance and Relief algorithms. The comparison is done over 19 datasets along with three different induction algorithms.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag 1998

Authors and Affiliations

  • Javier Lorenzo
    • 1
  • Mario Hernández
    • 1
  • Juan Méndez
    • 1
  1. 1.Dpto. de Informática y SistemasUniv. de Las Palmas de Gran CanariaLas PalmasSpain