Learning dynamic dependency network structure with time lag

Research Paper
  • 86 Downloads

Abstract

Characterizing and understanding the structure and the evolution of networks is an important problem for many different fields. While in the real-world networks, especially the spatial networks, the influence from one node to another tends to vary over both space and time due to the different space distances and propagation speeds between nodes. Thus the time lag plays an essential role in interpreting the temporal causal dependency among nodes and also brings a big challenge in network structure learning. However most of the previous researches aiming to learn the dynamic network structure only treat the time lag as a predefined constant, which may miss important information or include noisy information if the time lag is set too small or too large. In this paper, we propose a dynamic Bayesian model with adaptive lags (DBAL) which simultaneously integrates two usually separate tasks, i.e., learning the dynamic dependency network structure and estimating time lags, within one unified framework. Specifically, we propose a novel weight kernel approach for time series segmenting and sampling via leveraging samples from adjacent segments to avoid the sample scarcity. Besides, an effective Bayesian scheme cooperated with reversible jump Markov chain Monte Carlo (RJMCMC) and expectation propagation (EP) algorithm is proposed for parameter inference. Extensive empirical evaluations are conducted on both synthetic and two real-world datasets, and the results demonstrate that our proposed model is superior to the traditional methods in learning the network structure and the temporal dependency.

Keywords

dependency network time lag dynamic network 

Notes

Acknowledgements

This work was supported by National Science and Technology Support Plan (Grant No. 2014BAG01B02), National Natural Science Foundation of China (Grant No. 61572041), Joint Funds of the National Natural Science Foundation of China (Grant No. U1404604), and Beijing Natural Science Foundation (Grant No. 4152023).

References

  1. 1.
    Lebre S, Becq J, Devaux F, et al. Statistical inference of the time-varying structure of gene-regulation networks. BMC Syst Biol, 2010, 4: 130CrossRefGoogle Scholar
  2. 2.
    Goldenberg A, Moore A W. Bayes net graphs to understand co-authorship networks? In: Proceedings of the 3rd International Workshop on Link Discovery, New York, 2005. 1–8Google Scholar
  3. 3.
    Chen X, Liu Y, Liu H, et al. Learning spatial-temporal varying graphs with applications to climate data analysis. In: Proceedings of the 24th AAAI Conference on Artificial Intelligence, Atlanta, 2010Google Scholar
  4. 4.
    Dhurandhar A. Learning maximum lag for grouped graphical granger models. In: Proceedings of IEEE International Conference on Data Mining Workshops. Washington: IEEE Computer Society, 2010. 217–224Google Scholar
  5. 5.
    Barthélemy M. Spatial networks. Phys Rep, 2011, 499: 1–101MathSciNetCrossRefGoogle Scholar
  6. 6.
    Grzegorczyk M. A non-homogeneous dynamic Bayesian network with a hidden Markov model dependency structure among the temporal data points. Mach Learn, 2016, 102: 155–207MathSciNetCrossRefMATHGoogle Scholar
  7. 7.
    Liu Y, Kalagnanam J R, Johnsen O. Learning dynamic temporal graphs for oil-production equipment monitoring system. In: Proceedings of the 15th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Paris, 2009. 1225–1234Google Scholar
  8. 8.
    Song L, Kolar M, Xing E P. Time-varying dynamic Bayesian networks. In: Proceedings of Advances in Neural Information Processing Systems, Vancouver, 2009. 1732–1740Google Scholar
  9. 9.
    Dobigeon N, Tourneret J Y, Davy M. Joint segmentation of piecewise constant autoregressive processes by using a hierarchical model and a Bayesian sampling approach. IEEE Trans Signal Process, 2007, 55: 1251–1263MathSciNetCrossRefGoogle Scholar
  10. 10.
    Talih M, Hengartner N. Structural learning with time-varying components: tracking the cross-section of financial time series. J Roy Stat Soc B, 2005, 67: 321–341MathSciNetCrossRefMATHGoogle Scholar
  11. 11.
    Xuan X, Murphy K. Modeling changing dependency structure in multivariate time series. In: Proceedings of the 24th International Conference on Machine Learning, Omaha, 2007. 1055–1062Google Scholar
  12. 12.
    Knapp C, Carter G. The generalized correlation method for estimation of time delay. IEEE Trans Aoust Speech, 1976, 24: 320–327CrossRefGoogle Scholar
  13. 13.
    Green P J. Reversible jump Markov chain Monte Carlo computation and Bayesian model determination. Biometrika, 1995, 82: 711–732MathSciNetCrossRefMATHGoogle Scholar
  14. 14.
    Minka T P. Expectation propagation for approximate Bayesian inference. In: Proceedings of the 17th Conference on Uncertainty in Artificial Intelligence, Seattle, 2001. 362–369Google Scholar
  15. 15.
    Eaton D, Murphy K. Bayesian structure learning using dynamic programming and MCMC. arXiv:1206.5247Google Scholar
  16. 16.
    Guo F, Hanneke S, Fu W, et al. Recovering temporally rewiring networks: a model-based approach. In: Proceedings of the 24th International Conference on Machine Learning, Corvallis, 2007. 321–328Google Scholar
  17. 17.
    Husmeier D, Dondelinger F, Lebre S. Inter-time segment information sharing for non-homogeneous dynamic Bayesian networks. In: Proceedings of Conference and Workshop on Neural Information Processing Systems (NIPS), Vancouver, 2010. 901–909Google Scholar
  18. 18.
    Sakurai Y, Papadimitriou S, Faloutsos C. Braid: stream mining through group lag correlations. In: Proceedings of the 2005 ACM SIGMOD International Conference on Management of Data, Melbourne, 2005. 599–610Google Scholar
  19. 19.
    Tibshirani R. Regression shrinkage and selection via the lasso. J Roy Stat Soc B, 1996, 58: 267–288MathSciNetMATHGoogle Scholar
  20. 20.
    Dondelinger F, Lèbre S, Husmeier D. Non-homogeneous dynamic Bayesian networks with Bayesian regularization for inferring gene regulatory networks with gradually time-varying structure. Mach Learn, 2013, 90: 191–230MathSciNetCrossRefMATHGoogle Scholar
  21. 21.
    Ishwaran H, Rao J S. Spike and slab variable selection: frequentist and Bayesian strategies. Ann Stat, 2005, 33: 730–773MathSciNetCrossRefMATHGoogle Scholar
  22. 22.
    George E I, McCulloch R E. Approaches for Bayesian variable selection. Stat Sin, 1997, 7: 339–373MATHGoogle Scholar
  23. 23.
    Hernández-Lobato J M, Hernández-Lobato D, Suárez A. Expectation propagation in linear regression models with spike-and-slab priors. Mach Learn, 2015, 99: 437–487MathSciNetCrossRefMATHGoogle Scholar
  24. 24.
    Green P J, Hastie D I. Reversible jump MCMC. Genetics, 2009, 155: 1391–1403Google Scholar
  25. 25.
    Robert C P, Ryden T, Titterington D M. Bayesian inference in hidden Markov models through the reversible jump Markov chain Monte Carlo method. J Roy Stat Soc B, 2000, 62: 57–75MathSciNetCrossRefMATHGoogle Scholar
  26. 26.
    Gelman A, Rubin D B. Inference from iterative simulation using multiple sequences. Stat Sci, 1992, 7: 457–472CrossRefGoogle Scholar
  27. 27.
    Arnold A, Liu Y, Abe N. Temporal causal modeling with graphical granger methods. In: Proceedings of the 13th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Jose, 2007. 66–75Google Scholar
  28. 28.
    Murphy K. An introduction to graphical models. Rap Tech, 2001: 1–19Google Scholar
  29. 29.
    Dagum P, Galper A, Horvitz E. Dynamic network models for forecasting. In: Proceedings of the 8th International Conference on Uncertainty in Artificial Intelligence, Stanford, 1992. 41–48Google Scholar
  30. 30.
    Zhou X, Hong H, Xing X, et al. Mining dependencies considering time lag in spatio-temporal traffic data. In: Proceedings of International Conference on Web-Age Information Management, Qingdao, 2015. 285–296Google Scholar
  31. 31.
    Zhang R J, Co J, Lee S, et al. Carbonaceous aerosols in PM10 and pollution gases in winter in Beijing. J Environ Sci, 2007, 19: 564–571CrossRefGoogle Scholar

Copyright information

© Science China Press and Springer-Verlag GmbH Germany 2017

Authors and Affiliations

  1. 1.Key Laboratory of Machine Perception, Ministry of EducationPeking UniversityBeijingChina
  2. 2.Institute of ElectronicsChinese Academy of SciencesBeijingChina
  3. 3.School of Computer and Information EngineeringHenan Normal UniversityXinxiangChina

Personalised recommendations