Abstract Details
Activity Number:
|
19
|
Type:
|
Topic Contributed
|
Date/Time:
|
Sunday, August 9, 2015 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Section on Bayesian Statistical Science
|
Abstract #317129
|
View Presentation
|
Title:
|
Bayesian Variable Selection for Binary Outcomes in High-Dimensional Settings
|
Author(s):
|
Wenyi Wang* and Amir Nikooienejad and Valen E. Johnson
|
Companies:
|
MD Anderson Cancer Center and Texas A&M University and Texas A&M University
|
Keywords:
|
Non-local priors ;
Logistic models ;
high-dimensional variable selection
|
Abstract:
|
One of the important areas of high dimensional data analysis is variable selection where one might want to find the most contributing variables among the myriad of them. This important problem becomes more difficult when the number of observations is much less than the number of variables. For instance, in bioinformatics and in cancer genomics research where thousands of genes play the role of covariates with a small number of samples. During the past decade, many methods have been proposed such as ISIS and Adaptive LASSO. Recently Johnson and Russell proposed using non-local prior densities on Bayesian model parameters and showed it outperforms the commonly used penalized likelihood methods for continuous outcomes. We utilize these newly proposed non-local priors and introduce a new Bayesian method of variable selection named MOMLogit for the binary outcomes. We have tested this algorithm on simulation data, and demonstrated it performs well under different parameter settings. As our method outperforms the ISIS methods under common conditions, we expect our method to have a significant impact on related applications such as in bioinformatics and computational biology.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2015 program
|
For program information, contact the JSM Registration Department or phone (888) 231-3473.
For Professional Development information, 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.
2015 JSM Online Program Home
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.