Marv K comments on Variational Bayesian methods