Framework for Gibbs sampling of probabilistic models
Switch branches/tags
Nothing to show
Clone or download
Laura Dietz
Latest commit 020df7b Apr 13, 2015


bayes-stack: Parallel MCMC inference on graphical models

Bayes-stack is a framework for parallel probabilistic inference on graphical models. The framework provides infrastructure for easily implementing MCMC/Gibbs sampling methods capable of scaling to dozens of cores.

Along with the framework itself, several models using blocked Gibbs sampling are provided in network-topic-models/. See documentation: blob/stable/doc/usage.markdown