Estimation of multivariate linear functional relationships
多元线性函数关系的估计
Consistent estimation of the structural parameters in a multivariate linear functional relationship between two vector variables observed with error is considered. When the covariance matrix of the error terms contains unknown parameters, it is shown that the usual maximum likelihood approach generally yields inconsistent estimators. A natural modification of the likelihood equations leading to a consistent estimation procedure is proposed. The connexion with estimation in…