Estimation of parameters in heteroscedastic multiple regression model using leverage based near-neighbors.

In this study, we propose a Leverage Based Near-Neighbor (LBNN) method where prior information on the structure of the heteroscedastic error is not required. In the proposed LBNN method, weights are determined not from the near-neighbor values of the explanatory variables, but from their correspondi...

Full description

Bibliographic Details
Main Authors: Midi, Habshah, Rana, Sohel, Imon, A. H. M. R.
Format: Article
Language:English
English
Published: Asian Network for Scientific Information 2009
Online Access:http://psasir.upm.edu.my/id/eprint/14572/
http://psasir.upm.edu.my/id/eprint/14572/1/Estimation%20of%20parameters%20in%20heteroscedastic%20multiple%20regression%20model%20using%20leverage%20based%20near.pdf
_version_ 1848842431053692928
author Midi, Habshah
Rana, Sohel
Imon, A. H. M. R.
author_facet Midi, Habshah
Rana, Sohel
Imon, A. H. M. R.
author_sort Midi, Habshah
building UPM Institutional Repository
collection Online Access
description In this study, we propose a Leverage Based Near-Neighbor (LBNN) method where prior information on the structure of the heteroscedastic error is not required. In the proposed LBNN method, weights are determined not from the near-neighbor values of the explanatory variables, but from their corresponding leverage values so that it can be readily applied to a multiple regression model. Both the empirical and Monte Carlo simulation results show that the LBNN method offers substantial improvement over the existing methods. The LBNN has significantly reduced the standard errors of the estimates and also the standard errors of residuals for both simple and multiple linear regression models. Hence, the LBNN can be established as one reliable alternative approach to other existing methods that deal with heteroscedastic errors when the form of heteroscedasticity is unknown.
first_indexed 2025-11-15T07:59:01Z
format Article
id upm-14572
institution Universiti Putra Malaysia
institution_category Local University
language English
English
last_indexed 2025-11-15T07:59:01Z
publishDate 2009
publisher Asian Network for Scientific Information
recordtype eprints
repository_type Digital Repository
spelling upm-145722015-10-22T07:14:07Z http://psasir.upm.edu.my/id/eprint/14572/ Estimation of parameters in heteroscedastic multiple regression model using leverage based near-neighbors. Midi, Habshah Rana, Sohel Imon, A. H. M. R. In this study, we propose a Leverage Based Near-Neighbor (LBNN) method where prior information on the structure of the heteroscedastic error is not required. In the proposed LBNN method, weights are determined not from the near-neighbor values of the explanatory variables, but from their corresponding leverage values so that it can be readily applied to a multiple regression model. Both the empirical and Monte Carlo simulation results show that the LBNN method offers substantial improvement over the existing methods. The LBNN has significantly reduced the standard errors of the estimates and also the standard errors of residuals for both simple and multiple linear regression models. Hence, the LBNN can be established as one reliable alternative approach to other existing methods that deal with heteroscedastic errors when the form of heteroscedasticity is unknown. Asian Network for Scientific Information 2009 Article PeerReviewed application/pdf en http://psasir.upm.edu.my/id/eprint/14572/1/Estimation%20of%20parameters%20in%20heteroscedastic%20multiple%20regression%20model%20using%20leverage%20based%20near.pdf Midi, Habshah and Rana, Sohel and Imon, A. H. M. R. (2009) Estimation of parameters in heteroscedastic multiple regression model using leverage based near-neighbors. Journal of Applied Sciences, 9 (22). pp. 4013-4019. ISSN 1812-5654; ESSN: 1812-5662 http://scialert.net/jindex.php?issn=1812-5654 10.3923/jas.2009.4013.4019 English
spellingShingle Midi, Habshah
Rana, Sohel
Imon, A. H. M. R.
Estimation of parameters in heteroscedastic multiple regression model using leverage based near-neighbors.
title Estimation of parameters in heteroscedastic multiple regression model using leverage based near-neighbors.
title_full Estimation of parameters in heteroscedastic multiple regression model using leverage based near-neighbors.
title_fullStr Estimation of parameters in heteroscedastic multiple regression model using leverage based near-neighbors.
title_full_unstemmed Estimation of parameters in heteroscedastic multiple regression model using leverage based near-neighbors.
title_short Estimation of parameters in heteroscedastic multiple regression model using leverage based near-neighbors.
title_sort estimation of parameters in heteroscedastic multiple regression model using leverage based near-neighbors.
url http://psasir.upm.edu.my/id/eprint/14572/
http://psasir.upm.edu.my/id/eprint/14572/
http://psasir.upm.edu.my/id/eprint/14572/
http://psasir.upm.edu.my/id/eprint/14572/1/Estimation%20of%20parameters%20in%20heteroscedastic%20multiple%20regression%20model%20using%20leverage%20based%20near.pdf