Activity Number:
|
438
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 2, 2016 : 2:00 PM to 3:50 PM
|
Sponsor:
|
IMS
|
Abstract #319397
|
|
Title:
|
Sufficient Dimension Reduction with Missing Data
|
Author(s):
|
Qi Xia* and Yuexiao Dong and Chengyong Tang
|
Companies:
|
Temple University and Temple University and Temple University
|
Keywords:
|
Sufficient dimension reduction ;
Missing at random ;
Estimating equations ;
Nonparametric regression
|
Abstract:
|
Existing sufficient dimension reduction (SDR) methods typically consider cases with no missing data. In this study, we first show that biased and/or less efficient estimations may generally be the case for SDR methods when items in the observable data can be missing. We then consider a framework using estimating equation approach for SDR dealing with missing data. We show that consistent and more efficient estimators can be obtained by appropriately incorporating and adjusting for the impact due to data missingness. Simulations and data analyses are conducted for demonstrating the performance of the proposed framework.
|
Authors who are presenting talks have a * after their name.