|
Activity Number:
|
204
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Quality and Productivity
|
| Abstract - #309876 |
|
Title:
|
A Comparison of Model Combining Methods
|
|
Author(s):
|
Lihua Chen*+ and Panayotis Giannakouros
|
|
Companies:
|
The University of Toledo and University of Missouri-Kansas City
|
|
Address:
|
Mail Stop 942, Toledo, OH, 43606,
|
|
Keywords:
|
Bayesian Model Averaging ; Adaptive Regression by Mixing ; generalized linear models ; model combining
|
|
Abstract:
|
The present work compares a prediction-based model combining method to other model combining methods in generalized linear models. We implement Bayesian Model Averaging using a BIC approximation in a way that can accommodate interaction terms and therefore can be compared to prediction-based model combining across all parametric model settings in which prediction-based model combining is implemented. We also present results for several other weighting schemes for model combining.
|
