ABSTRACT A key challenge for modern Bayesian statistics is how to perform scalable
inference of posterior distributions. To address this challenge, variational Bayes (VB) …