scientific article; zbMATH DE number 4082772

From MaRDI portal
Publication:3811552





[For the entire collection see Zbl 0656.00023.] For a general (possibly, non-normal) multivariate linear model, under the plausibility of a redundant subset of parameters, James-Stein rule based shrinkage least squares estimation theory is considered. Relative risk- efficiencies of the classical least squares estimators and their shrinkage as well as preliminary test versions are studied (in the light of their asymptotic distributional risks), and it is shown that none of the adaptive versions dominates over the other, though each of them fares well relative to the classical version.




Cited in
(31)








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 Q3811552)