|
Activity Number:
|
430
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 2, 2016 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biometrics Section
|
|
Abstract #320479
|
View Presentation
|
|
Title:
|
Multivariate Copula-Based Regression Models for Longitudinal Data
|
|
Author(s):
|
Xin Tian* and Colin O. Wu
|
|
Companies:
|
National Heart, Lung, and Blood Institute and National Heart, Lung, and Blood Institute
|
|
Keywords:
|
longitudinal analysis ;
multivariate distribution ;
copula ;
nonparametric modeling ;
time-varying transformation models ;
dependence
|
|
Abstract:
|
In biomedical studies, it is an important challenge to estimate joint distribution of multivariate longitudinal outcome variables, where marginal outcome variables may have non-normal distributions that change with time and time-dependent covariates. We propose to use multivariate copulas to link marginal outcomes and model their dependence, while individual marginal distributions conditioning on longitudinal covariates may be estimated by a class of flexible time-varying transformation models. An application of these models is to jointly track multivariate risk factors over time. Our models and estimation method are demonstrated through an epidemiological study of childhood growth and health study.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2016 program
|
