2001•Fuzhou daxue xuebao. Ziran kexue banRequires access
A generalized estimator of the regression parameter in multivariate linear model
Zhou Zhong-mei
Open publisher page 0 citations
Abstract
A generalized estimator of the regression parameters is proposed in multivariate linear model. It is proved that the mean square error of β *(K) is less than the square error of the least square estimator when the design matrix is ill-conditioned. It is also proved that β *(K) is an admissible estimator. Finally, we discuss the properties of the mean square residual error of β *(K).