Variational Bayes (VB) [Beal, 2003] is a method for lower-bounding the evidence, or marginal likelihood, of a Bayesian latent variable model, which is then useful for model comparison. An approximation to the posterior distribution over model parameters is also obtained, and this can be used to obtain an