This is the program for the 2010 Joint Statistical Meetings in Vancouver, British Columbia.

Abstract Details

Activity Number: 353
Type: Contributed
Date/Time: Tuesday, August 3, 2010 : 10:30 AM to 12:20 PM
Sponsor: Section on Statistical Computing
Abstract - #306980
Title: Latent Class Profile Analysis with Dynamic Dirichlet Learning Process on Model Selection Problems
Author(s): Hsiu-Ching Chang* and Hwan Chung+
Companies: Michigan State University and Michigan State University
Address: , , 48823,
Keywords: Finite mixture models ; Latent class analysis ; Stage-sequential process ; Dirichlet process ; Polya urn scheme ; Stick-breaking prior

The Latent Class Profile Analysis (LCPA) provides a set of principles for systematic identification of two-level homogeneous subgroups of individuals. The LCPA models are most useful in substance use intervention research. Drug abuse researchers currently have several methods at their disposal for evaluating the fit of LCPA, depending on the software package being used. However, model selection in LCPA is a difficult statistical problem which has not been completely resolved. In our study, we propose a non-parametric two-level Dirichlet Process for LCPA models to select the number of classes and the number of class-profiles simultaneously. At the first level, the algorithm utilizes the learning process for latent classes, referred to as Dynamic Dirichlet Process Mixture Model. At the second level, the standard Dirichlet Process Mixture Model is applied to class-profile learning process.

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 2010 program

2010 JSM Online Program Home

For information, contact or phone (888) 231-3473.

If you have questions about the Continuing Education program, please contact the Education Department.