Mean square error matrix comparisons of estimators in linear regression (Q3725372)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 3956253
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Mean square error matrix comparisons of estimators in linear regression |
scientific article; zbMATH DE number 3956253 |
Statements
Mean square error matrix comparisons of estimators in linear regression (English)
0 references
1985
0 references
mean square error matrix
0 references
biased estimation
0 references
restricted least
0 references
squares estimator
0 references
pretest estimators
0 references
0 references
0 references
0 references
0.95974714
0 references
0.9563618
0 references
0 references
Estimators of the parameter vector in the linear regression model are often compared by means of the difference of their mean squared matrices. Criteria have been established in order to examine whether this difference is positive definite, i.e. one estimator is superior to the other. There are, however, several cases when the difference is singular and nonnegative definite. One such case is when the least squares estimator is compared with the equality restricted least squares estimator. This paper extends the former theory by including also the case of a singular difference. The results are applied to equality restricted and pretest estimators.
0 references