Comparing Scheffe-Type to Constant-Width Confidence Bounds in Regression (Q3314764)
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 3844848
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Comparing Scheffe-Type to Constant-Width Confidence Bounds in Regression |
scientific article; zbMATH DE number 3844848 |
Statements
Comparing Scheffe-Type to Constant-Width Confidence Bounds in Regression (English)
0 references
1983
0 references
constant-width confidence bounds
0 references
Scheffé-type bounds
0 references
simultaneous confidence bounds
0 references
multilinear regression
0 references
tables of relative efficiency ranges
0 references
ellipsoidal regions
0 references
0.8547612428665161
0 references
0.8138923645019531
0 references
0.8011211156845093
0 references
0.798079252243042
0 references
0.7963261008262634
0 references
Simultaneous confidence bounds for the multilinear regression model with intercept over ellipsoidal regions are considered. Conditions are given for when the Scheffé-type bound has optimal average accuracy compared to the constant-width bound in the sense of having smaller average width with respect to Lebesgue measure when coverage probabilities are equal. These conditions are the non-negativity of a function which depends on an angle which measures the size of the ellipsoidal region relative to the model design. Tables of relative efficiency ranges of the two types of bounds for simple linear regression over a finite interval in the balanced case are given. Some generalizations to error structures other than usual i.i.d. normal model are given.
0 references