JSM 2013 Home
Online Program Home
My Program

Abstract Details

Activity Number: 340
Type: Topic Contributed
Date/Time: Tuesday, August 6, 2013 : 10:30 AM to 12:20 PM
Sponsor: Mental Health Statistics Section
Abstract - #307661
Title: Model Selection Criteria Based on Computationally Intensive Estimators of the Expected Optimism
Author(s): Joseph Cavanaugh*+
Companies: University of Iowa
Keywords: Akaike information criterion ; bootstrap ; cross validation ; Mallows' conceptual predictive statistic ; Monte Carlo simulation ; variable selection
Abstract:

A model selection criterion is often formulated by constructing an approximately unbiased estimator of an expected discrepancy. The expected discrepancy reflects how well, on average, the fitted candidate model predicts "new" data generated under the true model. Its natural estimator, the estimated discrepancy, reflects how well the fitted candidate model predicts the data at hand. The expectation of the difference between these measures is known as the expected optimism. The selection criterion arises by adding an approximation of the expected optimism to the estimated discrepancy to correct for its negative bias. Classical approaches to obtaining this approximation often lead to simplistic corrections, derived using large-sample arguments, restrictive assumptions on the form of the candidate model, or both. The resulting corrections may fail to perform adequately in settings where the sample size is small or the requisite assumptions do not hold. In this talk, we propose computationally intensive approaches to approximating the expected optimism. We discuss the properties of the resulting criteria and illustrate their use in identifying biosignatures for treatment response.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2013 program




2013 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.

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

ASA Meetings Department  •  732 North Washington Street, Alexandria, VA 22314  •  (703) 684-1221  •  meetings@amstat.org
Copyright © American Statistical Association.