|
Activity Number:
|
310
- Advances and Novel Problems in Flexible Analysis of Clustered Data with Complex Structures
|
|
Type:
|
Invited
|
|
Date/Time:
|
Tuesday, August 1, 2017 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
|
Abstract #322016
|
|
|
Title:
|
Flexible Bayesian Additive Joint Models
|
|
Author(s):
|
Meike Köhler and Nikolaus Umlauf and Sonja Greven*
|
|
Companies:
|
Helmholtz Zentrum München and Universität Innsbruck and Ludwig-Maximilians-Universität München
|
|
Keywords:
|
joint models ;
biomarkers ;
longitudinal data ;
time-to-event data ;
P-splines ;
anisotropic smoothing
|
|
Abstract:
|
The joint modeling of longitudinal and time-to-event data is of growing importance to gain insights into the association between a longitudinal (bio)marker and an event process. We develop a general framework of flexible additive joint models that allows the specification of a variety of effects, such as smooth nonlinear, time-varying and random effects, in the longitudinal and survival parts of the models. Mean and variance of the longitudinal submodel as well as baseline hazard and potentially time-varying covariate and marker effects in the survival submodel can be specified using flexible additive predictors. By making use of Bayesian P-splines and functional random intercepts, we are in particular able to capture highly nonlinear subject-specific marker trajectories as well as a time-varying or non-linear association between the marker and the event process. The model is estimated within a Bayesian framework and implemented in the R-package bamlss.
|
Authors who are presenting talks have a * after their name.
