ON THE EFFECTS OF THE PRESENCE AND METHODS OF THE ELIMINATION HETEROSCEDASTICITY AND AUTOCORRELATION IN THE REGRESSION MODEL
The article presents the methodological and technical bases for the creation of regression models that adequately reflect reality. The focus is on methods of removing residual autocorrelation in models. Algorithms eliminating heteroscedasticity and autocorrelation of the regression model residuals:...
Main Authors: | , |
---|---|
Format: | Article |
Language: | Russian |
Published: |
Plekhanov Russian University of Economics
2016-08-01
|
Series: | Статистика и экономика |
Subjects: | |
Online Access: | https://statecon.rea.ru/jour/article/view/368 |
_version_ | 1797696872009170944 |
---|---|
author | Nina L. Timofeeva Artem I. Fedossev |
author_facet | Nina L. Timofeeva Artem I. Fedossev |
author_sort | Nina L. Timofeeva |
collection | DOAJ |
description | The article presents the methodological and technical bases for the creation of regression models that adequately reflect reality. The focus is on methods of removing residual autocorrelation in models. Algorithms eliminating heteroscedasticity and autocorrelation of the regression model residuals: reweighted least squares method, the method of Cochran-Orkutta are given. A model of "pure" regression is build, as well as to compare the effect on the dependent variable of the different explanatory variables when the latter are expressed in different units, a standardized form of the regression equation. The scheme of abatement techniques of heteroskedasticity and autocorrelation for the creation of regression models specific to the social and cultural sphere is developed. |
first_indexed | 2024-03-12T03:32:23Z |
format | Article |
id | doaj.art-055a64399795408a94866d793923e0fe |
institution | Directory Open Access Journal |
issn | 2500-3925 |
language | Russian |
last_indexed | 2024-03-12T03:32:23Z |
publishDate | 2016-08-01 |
publisher | Plekhanov Russian University of Economics |
record_format | Article |
series | Статистика и экономика |
spelling | doaj.art-055a64399795408a94866d793923e0fe2023-09-03T13:24:17ZrusPlekhanov Russian University of EconomicsСтатистика и экономика2500-39252016-08-010116717010.21686/2500-3925-2014-1-167-170367ON THE EFFECTS OF THE PRESENCE AND METHODS OF THE ELIMINATION HETEROSCEDASTICITY AND AUTOCORRELATION IN THE REGRESSION MODELNina L. Timofeeva0Artem I. Fedossev1МЭСИМЭСИThe article presents the methodological and technical bases for the creation of regression models that adequately reflect reality. The focus is on methods of removing residual autocorrelation in models. Algorithms eliminating heteroscedasticity and autocorrelation of the regression model residuals: reweighted least squares method, the method of Cochran-Orkutta are given. A model of "pure" regression is build, as well as to compare the effect on the dependent variable of the different explanatory variables when the latter are expressed in different units, a standardized form of the regression equation. The scheme of abatement techniques of heteroskedasticity and autocorrelation for the creation of regression models specific to the social and cultural sphere is developed.https://statecon.rea.ru/jour/article/view/368социально-культурная сферарегрессионные моделиэффективные управленческие решениягетероскедастичность моделиавтокорреляция остатков моделивзвешенный метод наименьших квадратовsocial and cultural sphereregression modelseffective management decisionsheteroscedasticity modelautocorrelation modelthe reweighted least squares method |
spellingShingle | Nina L. Timofeeva Artem I. Fedossev ON THE EFFECTS OF THE PRESENCE AND METHODS OF THE ELIMINATION HETEROSCEDASTICITY AND AUTOCORRELATION IN THE REGRESSION MODEL Статистика и экономика социально-культурная сфера регрессионные модели эффективные управленческие решения гетероскедастичность модели автокорреляция остатков модели взвешенный метод наименьших квадратов social and cultural sphere regression models effective management decisions heteroscedasticity model autocorrelation model the reweighted least squares method |
title | ON THE EFFECTS OF THE PRESENCE AND METHODS OF THE ELIMINATION HETEROSCEDASTICITY AND AUTOCORRELATION IN THE REGRESSION MODEL |
title_full | ON THE EFFECTS OF THE PRESENCE AND METHODS OF THE ELIMINATION HETEROSCEDASTICITY AND AUTOCORRELATION IN THE REGRESSION MODEL |
title_fullStr | ON THE EFFECTS OF THE PRESENCE AND METHODS OF THE ELIMINATION HETEROSCEDASTICITY AND AUTOCORRELATION IN THE REGRESSION MODEL |
title_full_unstemmed | ON THE EFFECTS OF THE PRESENCE AND METHODS OF THE ELIMINATION HETEROSCEDASTICITY AND AUTOCORRELATION IN THE REGRESSION MODEL |
title_short | ON THE EFFECTS OF THE PRESENCE AND METHODS OF THE ELIMINATION HETEROSCEDASTICITY AND AUTOCORRELATION IN THE REGRESSION MODEL |
title_sort | on the effects of the presence and methods of the elimination heteroscedasticity and autocorrelation in the regression model |
topic | социально-культурная сфера регрессионные модели эффективные управленческие решения гетероскедастичность модели автокорреляция остатков модели взвешенный метод наименьших квадратов social and cultural sphere regression models effective management decisions heteroscedasticity model autocorrelation model the reweighted least squares method |
url | https://statecon.rea.ru/jour/article/view/368 |
work_keys_str_mv | AT ninaltimofeeva ontheeffectsofthepresenceandmethodsoftheeliminationheteroscedasticityandautocorrelationintheregressionmodel AT artemifedossev ontheeffectsofthepresenceandmethodsoftheeliminationheteroscedasticityandautocorrelationintheregressionmodel |