Abstract Details
Activity Number:
|
62
|
Type:
|
Topic Contributed
|
Date/Time:
|
Sunday, August 3, 2014 : 4:00 PM to 5:50 PM
|
Sponsor:
|
Mental Health Statistics Section
|
Abstract #311599
|
|
Title:
|
Support Vector Hazard Regression for Predicting Event Times Subject to Censoring
|
Author(s):
|
Yuanjia Wang*+ and Donglin Zeng and Xiaoxi Liu
|
Companies:
|
Columbia University and University of North Carolina at Chapel Hill and University of North Carolina at Chapel Hill
|
Keywords:
|
Support vector machines ;
Risk bound ;
Predicting time-to-event ;
Kernel machines
|
Abstract:
|
Predicting dichotomous or continuous disease outcomes using powerful machine learning approaches has been studied extensively. However, how to learn prediction rules for time-to-event outcomes subject to right censoring has received little attention until very recently. Existing approaches rely on inverse probability weighting or rank-based methods, which are inefficient. We develop a novel support vector hazards regression (SVHR) approach to predict time-to-event outcomes based on the counting process and a series of support vector machines (SVM) for time-to-event outcomes among subjects at risk. Introducing counting processes to represent the time-to-event data leads to an intuitive connection of the method with SVM in standard supervised learning and hazard regression models in standard survival analysis. We demonstrate an interesting connection of the profiled empirical risk function with the Cox partial likelihood and formally show SVHR is optimal in discriminating covariate-specific hazard function from population average hazard function. We apply our method to analyze data from two real world studies to demonstrate superiority of SVHR in practical settings.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2014 program
|
2014 JSM Online Program Home
For information, contact jsm@amstat.org or phone (888) 231-3473.
If you have questions about the Professional Development program, please contact the Education Department.
The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.
Copyright © American Statistical Association.