Журнал «Современная Наука»

Russian (CIS)English (United Kingdom)
 MOSCOW +7(495)-142-86-81

Forecasting by extrapolation of a neural network error

Gusev A. L.  (Perm State National Research University)

Okunev A. A.  (Perm State National Research University)

In this paper, we describe a method for extrapolating a neural network error developed by the authors. The method is effective when observations are made for several time periods and for several territories (objects). The method of extrapolating the error of a neural network is based on the procedure for compressing the information space and the procedure for expanding the information space. The authors considered three methods of compressing the information space: compression by all indicators, by defining indicators and by the predicted indicator. The article considers four ways of extrapolating the error of a neural network. On a concrete example in work it is shown, that the method of extrapolation of a neural network error allows to reduce essentially the general average error of the forecast.

Keywords:indicator forecast, information space, neural network, compression procedure, extension procedure.
 

Read the full article …

Citation link:
Gusev A. L., Okunev A. A. Forecasting by extrapolation of a neural network error // Современная наука: актуальные проблемы теории и практики. Серия: Естественные и Технические Науки. -2018. -№02. -С. 32-37
LEGAL INFORMATION:
Reproduction of materials is permitted only for non-commercial purposes with reference to the original publication. Protected by the laws of the Russian Federation. Any violations of the law are prosecuted.
© ООО "Научные технологии"

Terms Publications

Issues of the journals (Archive)

ECONOMICS AND LAW
HUMANITIES
NATURAL AND TECHNICAL SCIENCES
COGNITION

Серия - Natural and Technical Sciences