JMboost
implements a boosting algorithm for fitting joint models to potentially
high-dimensional data. The algorithm provides the functionality to simultaneously
estimate a longitudinal and a shared predictor.
Details can be found here:
- JMboost manuscript (preliminary version currently under review).
Latest version from GitHub:
library("devtools")
install_github("mayrandy/JMboost")
library("JMboost")
To be able to use the install_github()
command, one needs to install devtools
first:
install.packages("devtools")