The Factored Frontier Algorithm for Approximate Inference in DBNs
This work addresses computational tractability issues in inference for DBNs, which is important for applications such as monitoring systems and traffic modeling, but it is incremental as it builds on existing methods like BK and LBP.
The paper tackled the problem of approximate inference in Dynamic Bayesian Networks (DBNs) by introducing the Factored Frontier (FF) algorithm, which simplifies computations by always working with factored distributions, and showed empirically that iterating loopy belief propagation (LBP) can improve accuracy over FF and the Boyen-Koller (BK) algorithm on real-world models like a water treatment plant and freeway traffic.
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.