Advances in Artificial Intelligence - IBERAMIA-SBIA 2006

Volume 4140 of the series Lecture Notes in Computer Science pp 299-308

A New Linear Dimensionality Reduction Technique Based on Chernoff Distance

  • Luis RuedaAffiliated withDepartment of Computer Science and Center for Biotechnology, University of Concepción
  • , Myriam HerreraAffiliated withDepartment and Institute of Informatics, National University of San Juan

* Final gross prices may vary according to local VAT.

Get Access


A new linear dimensionality reduction (LDR) technique for pattern classification and machine learning is presented, which, though linear, aims at maximizing the Chernoff distance in the transformed space. The corresponding two-class criterion, which is maximized via a gradient-based algorithm, is presented and initialization procedures are also discussed. Empirical results of this and traditional LDR approaches combined with two well-known classifiers, linear and quadratic, on synthetic and real-life data show that the proposed criterion outperforms the traditional schemes.