Sciact
  • EN
  • RU

Universal kernel-type estimators for the conditional variance in heteroscedastic models of nonparametric regression Full article

Journal Siberian Advances in Mathematics
ISSN: 1055-1344 , E-ISSN: 1934-8126
Output data Year: 2025, Volume: 35, Number: 4, Pages: 277-286 Pages count : 10 DOI: 10.1134/S1055134425040017
Tags nonparametric heteroscedastic regression, kernel estimator, conditional variance function, consistency, fixed design, random design, strongly dependent design
Authors Borisov I.S. 1 , Linke Yu.Yu. 1
Affiliations
1 Sobolev Institute of Mathematics

Funding (1)

1 Sobolev Institute of Mathematics FWNF-2024-0001

Abstract: The consistency of new universal kernel estimators for conditional variance function in a heteroscedastic nonparametric regression model has been proven. The new estimators are insensitive to the nature of the design dependence. For design that can be either fixed or random, only the following condition is used: the design points densely fill the domain of regression function. As a consequence, we consider the problem of constructing a confidence region for a regression function under the above-mentioned very general conditions on the design in terms of dense data.
Cite: Borisov I.S. , Linke Y.Y.
Universal kernel-type estimators for the conditional variance in heteroscedastic models of nonparametric regression
Siberian Advances in Mathematics. 2025. V.35. N4. P.277-286. DOI: 10.1134/S1055134425040017 Scopus
Dates:
Submitted: Sep 8, 2025
Accepted: Oct 29, 2025
Published online: Nov 19, 2025
Identifiers:
Scopus: 2-s2.0-105025421912
Citing: Пока нет цитирований
Altmetrics: