On the Generalization Ability of GRLVQ Networks

  • Authors:
  • Barbara Hammer;Marc Strickert;Thomas Villmann

  • Affiliations:
  • Department of Mathematics/Computer Science, Universität Osnabrück, Germany;Department of Mathematics/Computer Science, Universität Osnabrück, Germany;Clinique for Psychotherapy, Universität Leipzig, Germany

  • Venue:
  • Neural Processing Letters
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

We derive a generalization bound for prototype-based classifiers with adaptive metric. The bound depends on the margin of the classifier and is independent of the dimensionality of the data. It holds for classifiers based on the Euclidean metric extended by adaptive relevance terms. In particular, the result holds for relevance learning vector quantization (RLVQ) [4] and generalized relevance learning vector quantization (GRLVQ) [19].