On the Convergence of Stochastic Variational Inference in Bayesian Networks

Abstract

We highlight a pitfall when applying stochastic variational inference to general Bayesian networks. For global random variables approximated by an exponential family distribution, natural gradient steps, commonly starting from a unit length step size, are averaged to convergence. This useful insight into the scaling of initial step sizes is lost when the approximation factorizes across a general Bayesian network, and care must be taken to ensure practical convergence. We experimentally investigate how much of the baby (well-scaled steps) is thrown out with the bath water (exact gradients).

arXiv commentNIPS 2014 Workshop on Advances in Variational Inference. Montreal, Canada

References in corpus (1)

• One-class Collaborative Filtering with Random Graphs: Annotated Version

Cited by in corpus (2)

• Faster Stochastic Variational Inference using Proximal-Gradient Methods with General Divergence Functions
• A Factorial Mixture Prior for Compositional Deep Generative Models