Belief Propagation for Continuous State Spaces: Stochastic Message-Passing with Quantitative Guarantees
Nima Noorshams, Martin J. Wainwright.
Year: 2013, Volume: 14, Issue: 85, Pages: 2799−2835
Abstract
The sum-product or belief propagation (BP) algorithm is a widely used message-passing technique for computing approximate marginals in graphical models. We introduce a new technique, called stochastic orthogonal series message-passing (SOSMP), for computing the BP fixed point in models with continuous random variables. It is based on a deterministic approximation of the messages via orthogonal series basis expansion, and a stochastic estimation of the basis coefficients via Monte Carlo techniques and damped updates. We prove that the SOSMP iterates converge to a $\delta$-neighborhood of the unique BP fixed point for any tree-structured graph, and for any graphs with cycles in which the BP updates satisfy a contractivity condition. In addition, we demonstrate how to choose the number of basis coefficients as a function of the desired approximation accuracy $\delta$ and smoothness of the compatibility functions. We illustrate our theory with both simulated examples and in application to optical flow estimation.