Automatic Classification with Neural Networks Using New Decision Rule


Karimov A., Moharrami S.

INTERNATIONAL JOURNAL OF APPLIED MATHEMATICS & STATISTICS, vol.19, no.D10, pp.90-96, 2010 (ESCI) identifier identifier

  • Publication Type: Article / Article
  • Volume: 19 Issue: D10
  • Publication Date: 2010
  • Journal Name: INTERNATIONAL JOURNAL OF APPLIED MATHEMATICS & STATISTICS
  • Journal Indexes: Emerging Sources Citation Index (ESCI), Scopus
  • Page Numbers: pp.90-96
  • Azerbaijan State University of Economics (UNEC) Affiliated: Yes

Abstract

Neural networks Calculation is a subject that contains extensive domains in science. But no one of its important domains in deep perception neural networks is to the extent perception classification method in pattern recognition. Pattern recognition concepts have increasingly recognized as an important factor in the design of modern computerized systems. In this paper, The problem of automatic classification of the objects described by quantitative attributes is described and, the problem of redistribution of objects among classes is solved. Decision Rule was derived and given simultaneous transferring opportunity to the set of objects from the same class in another one by the minimum criterion of the sum of squares deviations. It is demanded to construct the optimal classification with use neural networks in New Decision Rule.