Model Selection for Generalized Linear Mixed Models

• Activity Number: 44
• Type: Contributed
• Date/Time: Sunday, August 1, 2010 : 2:00 PM to 3:50 PM
• Sponsor: WNAR
• Author(s): Rosanna Haut*+
• Companies: University of California, San Diego
• Keywords: Model Selection ; GLMM ; AIC

Abstract

This talk is about model selection for clustered data. A conditional Akaike Information, cAI, has been defined for linear mixed models when the desired inference is on cluster, rather than population, parameters. We extend this definition to cover the case of generalized linear mixed models. Using a second order Taylor expansion, we derive an asymptotically approximate estimate of cAI. This estimator is applied to a cancer data set. Alternatively, cAI can be estimated using a nonparametric bootstrap. We compare the two estimators in simulations.