Abstract
Dynamic Bayesian networks (DBNs) are probabilistic graphical models that have become a ubiquitous tool for compactly describing statistical relationships among a group of stochastic processes. A suite of elaborately designed inference algorithms makes it possible for intelligent systems to use a DBN to make inferences in uncertain conditions. Unfortunately, exact inference or even approximation in a DBN has been proved to be NP-hard and is generally computationally prohibitive. In this paper, we investigate a sliding window framework for approximate inference in DBNs to reduce the computational burden. By introducing a sliding window that moves forward as time progresses, inference at any time is restricted to a quite narrow region of the network. The main contributions to the sliding window framework include an exploration of its foundations, explication of how it operates, and the proposal of two strategies for adaptive window size selection. To make this framework available as an inference engine, the interface algorithm widely used in exact inference is then integrated with the framework for approximate inference in DBNs. After analyzing its computational complexity, further empirical work is presented to demonstrate the validity of the proposed algorithms.
Similar content being viewed by others
References
Murphy KP (2002) Dynamic Bayesian networks: representation, inference and learning. PhD Dissertation, University of California, Berkeley
Wang JS, Byrnes J, Valtorta M, Huhns M (2012) On the combination of logical and probabilistic models for information analysis. Appl Intell 36(2):472–497
Benferhat S, Boudjelida A, Tabia K, Drias H (2012) An intrusion detection and alert correlation approach based on revising probabilistic classifiers using expert knowledge. Appl Intell 36(4):520–540
Simona C, Koller P, Evsukoffb A (2012) DBN-based combinatorial resampling for articulated object tracking. In: Proceedings of UAI’12. Kaufmann, Los Altos, pp 237–246
Donat R, Leray P, Bouillaut L, Aknin P (2010) A dynamic Bayesian network to represent discrete duration models. Neurocomputing 73(4):570–577
Friedman N (2004) Inferring cellular networks using probabilistic graphical models. Science 303(5659):799–805
Zhang CL, Frias MA, Mele A, Ruggiu M, Eom T, Marney CB, Wang HD, Licatalosi DD, Fak JJ, Darnell RB (2011) Integrative modeling defines the nova splicing-regulatory network and its combinatorial controls. Science 329(5990):439–443
Bartels CD, Bilmes JA (2011) Creating non-minimal triangulations for use in inference in mixed stochastic/deterministic graphical models. Mach Learn 84(3):249–289
Yap GE, Tan AH, Pang HH (2008) Explaining inferences in Bayesian networks. Appl Intell 29(3):263–278
Koller D, Friedman N (2009) Probabilistic graphical models: principles and techniques. MIT Press, Cambridge
Smyth P, Heckerman D, Jordan MI (1997) Probabilistic independence networks for hidden Markov probability models. Neural Comput 9(2):227–269
Murphy K, Weiss Y (2001) The factored frontier algorithm for approximate inference in DBNs. In: Proceedings of UAI’01. Kaufmann, Los Altos, pp 378–385
Cano A, Gémez-Olmedo M, Moral S (2011) Approximate inference in Bayesian networks using binary probability trees. Int J Approx Reason 52(1):49–62
Palaniappan SK, Akshay S, Liu B, Genest B, Thiagarajan PS (2012) Hybrid factored frontier algorithm for dynamic Bayesian networks with a biopathways application. IEEE/ACM Trans Comput Biol Bioinform 9(5):1352–1365
Cooper GF (1990) The computational complexity of probabilistic inference using Bayesian belief networks. Artif Intell 42(2):393–405
Dagum P, Luby M (1993) Approximating probabilistic inference in Bayesian belief networks is NP-hard. Artif Intell 60(1):141–153
Boyen X, Koller D (1998) Tractable inference for complex stochastic processes. In: Proceedings of UAI’98, Madison, Wisconsin. Kaufmann, Los Altos, pp 33–42
Kjærulff U (1992) A computational scheme for reasoning in dynamic probabilistic networks. In: Proceedings of UAI’92, Stanford, CA. Kaufmann, Los Altos, pp 121–129
Kjærulff U (1995) dHugin: a computational system for dynamic time-sliced Bayesian networks. Int J Forecast 11(1):89–111
Cho HC (2006) Dynamic Bayesian networks for online stochastic modeling. PhD thesis, University of Nevada, Reno, NV
Qi Y (2004) Extending expectation propagation for graphical models. PhD thesis, Massachusetts Institute of Technology, MA
Wei T (2007) Expectation propagation algorithm for Bayesian inference in dynamic systems. MS thesis, University of Texas, San Antonio, TX
Cho HC, Lee KS, Fadali MS (2009) Online learning algorithm for dynamic Bayesian networks for nonstationary signal processing. Int J Innov Comput Inf Control 5(4):1037–1041
Knox WB, Mengshoel OJ (2009) Diagnosis and reconfiguration using Bayesian networks: an electrical power system case study. In: Proceedings of IJCAI’09 workshop self-* and autonomous systems: reasoning and integration challenges (SAS-09), pp 67–74
Chen HY (2011) Study on inference of structure-variable dynamic Bayesian networks. PhD thesis, Northwestern Polytechnical University, Xi’an, China
Russell SJ, Norvig P, Canny JF, Malik JM, Edwards DD (1995) Artificial intelligence: a modern approach. Pearson Education, Upper Saddle River
Pavlovic VI (1999) Dynamic Bayesian networks for information fusion with applications to human-computer interfaces. PhD thesis, University of Illinois, Urbana-Champaign, IL
Meyn SSP, Richard LT (2009) Markov chains and stochastic stability. Cambridge University Press, New York
Chatterjee S, Russell S (2010) Why are DBNs sparse? In: Proceedings of AISTATS’10, pp 81–88
Lauritzen SL, Spiegelhalter D (1988) Local computations with probabilities on graphical structures and their application to expert systems (with discussion). J R Stat Soc B 50:157–224
Jensen FV, Nielsen TD (2007) Bayesian networks and decision graphs, 2nd edn. Springer, New York
Barber D (2012) Bayesian reasoning and machine learning. Cambridge University Press, New York
Acknowledgements
This work was supported by the National Nature Science Foundation of China (NSFC) under Grant 60774064 and the doctoral Fund of Ministry of Education of China under Grant 20116102110026. The authors thank the anonymous reviews for their insightful comments and constructive suggestions to improve this paper. We would also like to thank Wen Zengkui for his valuable advice on the structure of the paper and his remarkable work of spelling checking.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Gao, XG., Mei, JF., Chen, HY. et al. Approximate inference for dynamic Bayesian networks: sliding window approach. Appl Intell 40, 575–591 (2014). https://doi.org/10.1007/s10489-013-0486-9
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10489-013-0486-9