Feature evaluation and sub-class determination through functional mapping

  • Authors:
  • M Shridhar;A Badreldin

  • Affiliations:
  • Department of Electrical Engineering, University of Windsor, Windsor, Ontario, Canada N9B 3P4;Department of Electrical Engineering, University of Windsor, Windsor, Ontario, Canada N9B 3P4

  • Venue:
  • Pattern Recognition Letters
  • Year:
  • 1985

Quantified Score

Hi-index 0.10

Visualization

Abstract

This paper introduces a functional mapping technique for evaluating multidimensional features and determining subclasses for pattern recognition problems. Several examples from character recognition and speaker recognition systems are used to illustrate the effectiveness of the proposed technique. The method is also extended to analysis of the convergence properties of the covariance matrices.