|
Activity Number:
|
130
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, August 4, 2008 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #302031 |
|
Title:
|
Model Selection for Clustered Outcome Common Predictor Effect Models
|
|
Author(s):
|
Juan Jia*+ and Robert E. Weiss
|
|
Companies:
|
University of California, Los Angeles and University of California, Los Angeles
|
|
Address:
|
Department of Biostatistics , Los Angeles, CA, 90095,
|
|
Keywords:
|
Multivariate longitudinal data ; Common effects ; Leaps and bounds
|
|
Abstract:
|
Model selection is the problem of identifying the best models among a large number of competing models. Many model selection studies focus on variable selection. The Clustered Outcome common Predictor Effect (COPE) model clusters outcomes that respond in like fashion to all predictors. As the number of outcomes increases, the number of ways to cluster the outcomes increases rapidly. A leaps and bounds style algorithm is developed to efficiently seek a set of best COPE models. We apply the proposed method to multivariate longitudinal data from children of HIV+ parents. We seek the best ways to cluster nine psychometric sub-scales of the Basic Symptom Inventory and we use age, time in study, gender, season and parental drug use as predictors.
|
