Hyperspherical neighbourhoods and pattern recognition using neural networks

https://doi.org/10.1016/0020-0190(96)00086-5Get rights and content

Abstract

A definition of maximal neighbourhoods in the form of hyperspheres around points in metric spaces is presented. An approximate algorithm to compute the radii of these hyperspheres is described. An exact algorithm to solve the same problem is described next. This exact algorithm is shown to be capable of handling two simple generalisations of the problem scenario. Usefulness of hyperspherical neighbourhoods around points representing patterns in a neural network model is demonstrated by means of numerical simulations.

References (13)

  • T. Ohya et al.

    Inform. Process. Lett.

    (1984)
  • G. Athithan
  • G. Athithan, in...
  • F. Aurenhammer

    ACM Comput. Surveys

    (1991)
  • A. Bowyer

    Comput. J.

    (1981)
  • J.H. Conway et al.

    Sphere Packings, Lattices and Groups

    (1988)
There are more references available in the full text version of this article.

Cited by (0)

View full text