scientific article; zbMATH DE number 3934245

From MaRDI portal
Publication:3706358





[For the entire collection see Zbl 0572.00016.] In regression, this article establishes nonlinear versions of the Gauss- Markov theorem in prediction as well as in estimation, which are presented in a conditional (stronger) form as well as in an unconditional form. In addition, it is shown that an estimator of the form \(\bar y=C(y)y\) with \(C(y)X=I\) is a GLSE with an estimated covariance matrix, and that under normality, the MLE is a GLSE and always unbiased.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3706358)