JSM 2011 Online Program

The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.

Abstract Details

Activity Number: 256
Type: Contributed
Date/Time: Monday, August 1, 2011 : 2:00 PM to 3:50 PM
Sponsor: Section on Statistical Learning and Data Mining
Abstract - #303057
Title: Exploring Genetic Risk for Breast Cancer Using an Ensemble of Tree-Based Classifier
Author(s): Bethany Wolf*+ and Elizabeth Hill and Elizabeth H. Slate and Carola Neumann and Emily Kistner-Griffin
Companies: Medical University of South Carolina and Medical University of South Carolina and Medical University of South Carolina and Medical University of South Carolina and Medical University of South Carolina
Address: 135 Cannon Street, Charleston, SC, 29425,
Keywords: esemble classifiers ; binary predictors ; predictor interactions
Abstract:

The predictive accuracy of tree-based classifiers (learners) can be improved by using an ensemble of learners to predict an observation's class. Ensembles allow averaging across weak learners (unbiased classifiers that are highly variable) resulting in an unbiased aggregated learner with reduced variability. Some ensemble methods also provide measures of predictor importance, allowing scientists to discover potential biological markers predictive of disease. For example, single nucleotide polymorphisms (SNPs) are thought to alter risk of developing disease or prognosis once disease occurs. Measures of predictor importance from an ensemble enable ranking of SNPs and SNP interactions according to association with disease outcome. We examine the performance of several ensemble methods in simulation studies for predictive capability and for ability to correctly identify binary predictors and predictor interactions associated with a binary outcome. We apply these methods to a subset of data from the Cancer Genetic Markers of Susceptibility (CGEMS) Breast Cancer Scan which includes SNPs from select genes and explore associations between the SNPs on these genes and breast cancer.


The address information is for the authors that have a + after their name.
Authors who are presenting talks have a * after their name.

Back to the full JSM 2011 program




2011 JSM Online Program Home

For information, contact jsm@amstat.org or phone (888) 231-3473.

If you have questions about the Continuing Education program, please contact the Education Department.