|
Activity Number:
|
508
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 10, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
| Abstract - #306775 |
|
Title:
|
Multivariate Robust Regression Based on General Depth Function
|
|
Author(s):
|
Weihua Zhou*+
|
|
Companies:
|
The University of North Carolina at Charlotte
|
|
Address:
|
Department of Mathematics and Statistics, Charlotte, NC, 28223,
|
|
Keywords:
|
multivariate regression ; robust ; depth function ; influence function ; breakdown point
|
|
Abstract:
|
A new estimator of the regression parameters is introduced in a multivariate regression model. The affine equivariant estimate is based on the general depth-weighted mean and scatter estimators. The influence function, finite breakdown and asymptotic theorem are developed to consider robustness and limiting efficiencies of this new estimate. The estimate is shown to be fisher consistent with a limiting multivariate normal distribution. The influence function, as a function of the length of the contaminated vector, is shown to be bounded in elliptic cases. The new estimate is highly efficient in the multivariate normal case and it is also highly robust. Simulations are used to consider finite sample efficiencies with similar results.
|
