No label defined (Q3709633)
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 3938276
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | No label defined |
scientific article; zbMATH DE number 3938276 |
Statements
1983
0 references
second order availability
0 references
positive solutions
0 references
best equivariant estimator
0 references
loss functions
0 references
second order properties of estimators
0 references
0.8765074014663696
0 references
0.7569343447685242
0 references
0.7437366247177124
0 references
[For the entire collection see Zbl 0509.00020.] As it is widely known the best equivariant estimator X of the unknown mean vector \(\theta \in R^ S\) of a normal family N(\(\theta\),E), \(E=(\delta_{ij})\), is admissible in \(R^ S\) only for \(S=1,2\). Estimators \(\theta\) (X) were exhibited in \textit{C. Stein}, Proc. 3rd Berkeley Sympos. math. Statist. Probability 1, 197-206 (1956; Zbl 0073.356), such that for \(S>2\) \[ (1)\quad E_{\theta}| \theta (X)-\theta |^ 2=E_{\theta}| X-\theta |^ 2-V(\theta),\quad V(\theta)>0,\quad \theta \in R^ S. \] It is but natural to inquire what is the set of the functions V(\(\theta)\) available in (1). Here we examine the problem in an asymptotic setup for which the covariance matrix E is substituted by \(\gamma\) E for \(\gamma\) \(\to 0\), while the relation (1) is appropriately replaced by \[ (2)\quad E_{\theta}| \theta (X)-\theta |^ 2=E_{\theta}| X-\theta |^ 2-\gamma^ 2V(\theta)+o(\gamma^ 2),\quad \gamma \to 0, \] holding uniformly with respect to any compact subset of the given parametric set \(\Theta\). Such an approach clarifies the nature of the restraint imposed on V(\(\theta)\) by the relation (2), allows a substantial extension of the class of loss functions involved in the analysis and furthermore helps to make more transparent the intrinsic connection between the second order properties of estimators and the theory of partial elliptic equations.
0 references