Cuts in Bayesian graphical models

Abstract The cut function defined by the OpenBUGS software is described as a “valve” that prevents feedback in Bayesian graphical models. It is shown that the MCMC algorithm applied by OpenBUGS in the presence of a cut function does not converge to a well-defined limiting distribution. However, it may be improved by using tempered transitions. The cut algorithm is compared with multiple imputation as a gold standard in a simple example. Keywords Bayesian inference · Cutting feedback · Multiple imputation

• Type: Journal
• Year: 2015
• Where: SAC
• Authors: Martyn Plummer
• Keywords: Applied Computing, SAC 2015