The Factored Frontier Algorithm for Approximate Inference in DBNs

TL;DR

The paper introduces the Factored Frontier algorithm for approximate inference in Dynamic Bayesian Networks, demonstrating its equivalence to loopy belief propagation and showing empirical improvements over related methods.

Contribution

It presents the FF algorithm as a tractable approximation method and establishes its theoretical connection to loopy belief propagation in DBNs.

Findings

FF is equivalent to one iteration of loopy belief propagation.

Iterating LBP improves accuracy over FF and Boyen-Koller.

Empirical results on water treatment and traffic models validate the approach.

Abstract

The Factored Frontier (FF) algorithm is a simple approximate inferencealgorithm for Dynamic Bayesian Networks (DBNs). It is very similar tothe fully factorized version of the Boyen-Koller (BK) algorithm, butinstead of doing an exact update at every step followed bymarginalisation (projection), it always works with factoreddistributions. Hence it can be applied to models for which the exactupdate step is intractable. We show that FF is equivalent to (oneiteration of) loopy belief propagation (LBP) on the original DBN, andthat BK is equivalent (to one iteration of) LBP on a DBN where wecluster some of the nodes. We then show empirically that byiterating, LBP can improve on the accuracy of both FF and BK. Wecompare these algorithms on two real-world DBNs: the first is a modelof a water treatment plant, and the second is a coupled HMM, used tomodel freeway traffic.