Iterative Refinement of the Approximate Posterior for Directed Belief Networks

Open in new window