Semiparametric reconstruction of the density function based on a generalized lambda-distribution in the problem of identification of regression models

V. I. Denisov, V. S. Timofeev, E. A. Khailenko

Novosibirsk State Technical University, 20 Karl Marx av., 630073 Novosibirsk

Abstract: The problem of estimating the parameters of regression models is considered. We study the method of the adaptive estimation of the parameters of regression models with the use of the semiparametric approach to the estimation of the density distribution function of random errors. The accuracy of the estimation of the parameters of the regression dependencies of this method is compared with the results obtained by the adaptive method on the basis of the generalized lambda-distribution developed earlier by the authors.

Keywords: regression dependency, adaptive estimation, nuclear estimate, generalized lambda-distribution, maximum likelihood method, identification of a distribution.

UDC: 519.242.5

Received: 27.05.2014

 English version: Journal of Applied and Industrial Mathematics, 2014, 8:4, 521–527

Bibliographic databases:

•