|
Activity Number:
|
605
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 6, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
| Abstract - #304797 |
|
Title:
|
Asymptotic Expansions for Dimension Reduction Methods with Application to Bias Correction
|
|
Author(s):
|
Zhou Yu*+ and Bing Li and Li-Xing Zhu
|
|
Companies:
|
East China Normal University and Penn State University and Hong Kong Baptist University
|
|
Address:
|
429 Oakwood Avenue, State College, PA, 16803,
|
|
Keywords:
|
Asymptotic expansion ; Bias correction ; Sliced inverse regression ; Sliced average variance estimation ; Principle Hessian direction
|
|
Abstract:
|
We in this paper develop the high order asymptotic expansion of classical eigen-decomposition type sufficient dimension reduction methods, such as Sliced Inverse Regression, Sliced Average Variance Estimation and Principle Hessian Directions, etc. Moreover, we propose a general bias correction strategy for existing dimension reduction estimator in order to removing the leading term from the asymptotic bias. The efficiency of our proposed bias corrected estimators is illustrated by comprehensive simulated comparisons.
|
