Note on Generalized Least Squares with Stepwise Estimation Procedure

Abstract

일반화한 선형회귀모형에서 새로운 변수가 첨가되었을때 회귀계수의 일반화한 최소자승 추정치를 단계적 추정치(stepwise estimator)로 나타내었고, 첨가된 새변수에 대한 검정기준을 제시하였다.

The generalized least squares (GLS) method for estimating the parameters of a linear regression model y=Xβ+e with E［e］=0 and E［ee'］=Vσ²is commonly used. When the further regressors x??s are introduced for the given GLS model y=Xβ+e, the GLS estimators for the parameters are obtained by direct calculation from the normal equation. While the GLS estimators may also be obtained by the stepwise estimation procedure.

The generalized least squares (GLS) method for estimating the parameters of a linear regression model y=Xβ+e with E［e］=0 and E［ee'］=Vσ²is commonly used. When the further regressors x??s are introduced for the given GLS model y=Xβ+e, the GLS estimators for the parameters are obtained by direct calculation from the normal equation. While the GLS estimators may also be obtained by the stepwise estimation procedure.