Abstract
In this paper we derive the change-of-variance function of M-estimators of scale under general contamination, thereby extending the formula in Hampel et al. (1986). We say that an M-estimator is B-robust if its influence function is bounded, and we call it V-robust if its change-of-variance function is bounded from above. It is shown, for a natural class of M-estimators, that the general notion of V-robustness still implies B-robustness. Several classes of M-estimators are studied closely, as well as some typical examples and their interpretation.
Original language | English (US) |
---|---|
Pages (from-to) | 69-80 |
Number of pages | 12 |
Journal | Journal of Computational and Applied Mathematics |
Volume | 64 |
Issue number | 1-2 |
DOIs | |
State | Published - Nov 30 1995 |
Externally published | Yes |
Keywords
- B-robustness
- Change-of-variance function
- Influence function
- V-robustness
ASJC Scopus subject areas
- Computational Mathematics
- Applied Mathematics