Abstract Details
Activity Number:
|
367
|
Type:
|
Invited
|
Date/Time:
|
Tuesday, August 5, 2014 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Section on Bayesian Statistical Science
|
Abstract #310926
|
View Presentation
|
Title:
|
Feature Allocation Models for Phenotype Discovery
|
Author(s):
|
Finale Doshi-Velez*+
|
Companies:
|
Harvard
|
Keywords:
|
feature allocation model ;
phenotype discovery ;
Indian Buffet Process
|
Abstract:
|
Clustering models are frequently used to find similar subgroups in a larger population. In medicine, clustering models are frequently used subtype patients into groups with similar disease characteristics, known as phenotypes. However, patients with complex conditions may express not one, but several phenotypes. Feature allocation models provide an approach to model this one-to-many relationship.
Here, we describe two feature allocation models for phenotype discovery, both built from the Indian Buffet Process. The first allows one to specify a prior distribution over the number of features independently for each observation. The second uses available knowledge of relationships between dimensions to produce sparser phenotype descriptions. We show that being able to incorporate covariates and other expert knowledge improves interpretability without sacrificing predictive power, a feature important to scientists and clinicians.
|
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.