|
Activity Number:
|
135
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
IMS
|
| Abstract - #309790 |
|
Title:
|
Bayesian Hidden Markov Models for Alcoholism Data
|
|
Author(s):
|
Kenneth Shirley*+ and Dylan Small and Kevin G. Lynch
|
|
Companies:
|
University of Pennsylvania and University of Pennsylvania and University of Pennsylvania
|
|
Address:
|
400 Jon M. Huntsman Hall, 3730 Walnut Street, Philadelphia, PA, 19104,
|
|
Keywords:
|
Hidden Markov Models ; Alcoholism ; Longitudinal Data ; Gibbs Sampler ; Forward-Backward Recursion
|
|
Abstract:
|
Alcoholism data typically consists of daily drink counts for recovering alcoholics, and covariates, some of which are time-varying and are also measured on a daily basis. Hidden Markov Models do a good job of capturing different drinking behaviors over time for this type of data. This paper incorporates covariates into an HMM for alcoholism data to model relationships between covariates and the outcome. Bayesian methods are used to fit the models, and predictions from the model are compared to those from other models. The clinical interpretability of estimates from the HMM is emphasized.
|
