|
Activity Number:
|
362
|
|
Type:
|
Invited
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Teaching Statistics in the Health Sciences
|
| Abstract - #307908 |
|
Title:
|
Conditional AIC for Nonlinear Mixed Effects Models
|
|
Author(s):
|
Florin Vaida*+
|
|
Companies:
|
University of California, San Diego
|
|
Address:
|
Division of Biostatistics, 9600 Gilman Drive, MC-0717, La Jolla, CA, 92093,
|
|
Keywords:
|
effective degrees of freedom ; model selection
|
|
Abstract:
|
In this paper we propose a model selection criterion for nonlinear and generalized linear mixed-effects model (NLME, GLME). The conditional AIC of Vaida and Blanchard (2005) is extended to NLME, using an appropriate definition of the effective degrees of freedom of the model, rho. This rho was proposed for GLME by Lu, Hodges and Carlin (2006). The criterion approximates the conditional Akaike information and the goodness of the approximation depends on the degree of non-linearity of the model. The conditional AIC is useful when the purpose of the model is subject-specific rather than population-level prediction. We use the criterion for model selection in the analysis of data from an ongoing international study of acute HIV infection.
|
- 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 2007 program |
