MODELING DEPENDENCIES «MULTIDIMEN¬SIONAL INPUT–OUTPUT» FOR AUTOMATION OF CONTROL PROCESSES UNDER UNCERTAINTY

Authors

  • Л. І. Лєві

DOI:

https://doi.org/10.31210/visnyk2015.04.22

Keywords:

multivariate dependence, multiple regression, fuzzy logic inference, soft computing, membership functions, (linguistic) variables

Abstract

In work considered technology allows to build multivariate dependence with continuous output by combining the advantages of soft computing and regression analysis, given the opportunity, the definition of importance of input variables and their necessary interactions. However, when modeling objects with continuous output when a sufficient accuracy of the determination of a precise value of the output value is necessary, the identification of the parameters of fuzzy regression equations using the least squares method and parameters of membership functions by statistical processing of expert information is not sufficient to provide the desired accuracy. It requires configuration on the training set of a fuzzy regression model in accordance with the testing sample.

Published

2015-12-25

How to Cite

Лєві, Л. І. (2015). MODELING DEPENDENCIES «MULTIDIMEN¬SIONAL INPUT–OUTPUT» FOR AUTOMATION OF CONTROL PROCESSES UNDER UNCERTAINTY. Scientific Progress & Innovations, (4), 86–90. https://doi.org/10.31210/visnyk2015.04.22

Issue

Section

ТЕХНІЧНІ НАУКИ