About one approach to the minimization of the errors of the tutoring of the neuron networks

  • Authors:

  • Alexander P. Grinko;Michal M. Karpuk

  • Affiliations:

  • Koszalin Technical University, Koszalin, Poland;Koszalin Technical University, Koszalin, Poland

  • Venue:
  • ICCMSE '03 Proceedings of the international conference on Computational methods in sciences and engineering
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

The feasibility of function of errors with fractional exponent for solving of a problem of optimization and tutoring of neural networks was theoretically explored. The analytical expressions for estimation of parameters of the models or weight factors were obtained. The algorithms were designed and the numerical experiment on actual economic datas was held, where the efficiency of an offered procedure is shown.