|
Activity Number:
|
544
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 3, 2016 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistical Learning and Data Science
|
|
Abstract #318852
|
|
|
Title:
|
An approximate L0-based variable selection method for high dimensional data
|
|
Author(s):
|
Zhihua Sun* and Gang Li
|
|
Companies:
|
Ocean University of China and University of California at Los Angeles
|
|
Keywords:
|
variable selection ;
high-dimensional censored data ;
linear regression
|
|
Abstract:
|
Feature selection from the high-dimensional censored data is a fundamental and challenging approach in recent years. In this talk, we present a novel algorithm for the high-dimensional right-censored data in linear regression. The proposed algorithm can identify effective nonzero features than the traditional variable selection procedures via extensive simulations. The performance of our method is demonstrated by medical studies.
|
Authors who are presenting talks have a * after their name.
