zhenkewu/doubletree: Nested Latent Class Models for Domain-Adaptive and Semisupervised Learning of Cause-of-Deaths using Verbal Autopsy

This package implements variational algorithm to probablistically assign cause-of-death using multivariate binary data (verbal autopsy). The model is for semi-supervised learning of unobserved causes of deaths for subjects in a target domain using data from training domains. The model is equipped with two components of the prior: 1) a tree-structured prior that promotes similarity in the conditional dependence structure amongst domains, and 2) a tree-structured prior that promotes similarity in the class response probability profiles amongst causes.

install.packages("remotes")
remotes::install_github("zhenkewu/doubletree")