Robust alternatives of the least squares estimator |
| |
| Authors: | Kurt Hoffmann |
| |
| Institution: | Sektion Mathematik , Humboldt-Universit?t , Unter den Linden 6 Postfach 1297, Berlin, DDR-1086G.D.R |
| |
| Abstract: | The purpose of this paper consists in deriving estimators which are less sensitive than the least squares estimator, when the assumption that the expectation vector lies in a certain linear subspace is violated. The obtained robust estimators are convex combinations of the least squares estimator and of the random vector Y. |
| |
| Keywords: | |
|
|
