|
Activity Number:
|
581
- Recent Advances in High-Dimensional Data Estimation and Prediction
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2018 : 2:00 PM to 3:50 AM
|
|
Sponsor:
|
Section on Statistical Computing
|
|
Abstract #330097
|
Presentation
|
|
Title:
|
Sure Independent Screening for Ultra-high Longitudinal Data
|
|
Author(s):
|
Yafei Zhang* and Pang Du
|
|
Companies:
|
Virginia Tech and Virginia Tech
|
|
Keywords:
|
longitudinal data;
graphical modeling;
linear mixed effect model
|
|
Abstract:
|
Ultrahigh-dimensional variable selection plays an increasingly important role in contemporary scientific discoveries and statistical research. Independence screening is powerful for variable selection when the number of variables is massive. Commonly used independence screening methods are based on single replicate data and are not applicable to longitudinal data. This motivates us to propose a Sure Independence Screening (SIS) procedure to reduce the dimension from ultrahigh to a scale similar to but smaller than the sample size. We provide two types of SIS, and their iterative extensions (iSIS) are also proposed to enhance the finite sample performance. We give conditions under which sure screening is possible and derive an upper bound on the number of selected variables. Moreover, we spell out the conditions under which SIS yields model selection consistency and possesses the sure independence screening property when a data-driven conditioning set is used. The proposed procedures are assessed by simulations and an application of them to a study on systemic lupus erythematosus illustrates the practical use of these procedures.
|
Authors who are presenting talks have a * after their name.
