Clinical Time Series Prediction with a Hierarchical Dynamical System

  • Zitao Liu
  • Milos Hauskrecht
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7885)


In this work we develop and test a novel hierarchical framework for modeling and learning multivariate clinical time series data. Our framework combines two modeling approaches: Linear Dynamical Systems (LDS) and Gaussian Processes (GP), and is capable to model and work with time series of varied length and with irregularly sampled observations. We test our framework on the problem of learning clinical time series data from the complete blood count panel, and show that our framework outperforms alternative time series models in terms of its predictive accuracy.


Time Series Gaussian Processes Linear Dynamical System 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Combi, C., Keravnou-Papailiou, E., Shahar, Y.: Temporal information systems in medicine. Springer Publishing Company (2010) (Incorporated)Google Scholar
  2. 2.
    Kalman, R.: Mathematical description of linear dynamical systems. Journal of the Society for Industrial & Applied Mathematics, Series A: Control 1, 152–192 (1963)MathSciNetzbMATHCrossRefGoogle Scholar
  3. 3.
    Rasmussen, C., Williams, C.: Gaussian processes for machine learning, vol. 1. MIT Press, Cambridge (2006)zbMATHGoogle Scholar
  4. 4.
    Rasmussen, C., Kuss, M., et al.: Gaussian processes in reinforcement learning. Advances in Neural Information Processing Systems 16 (2004)Google Scholar
  5. 5.
    Deisenroth, M., Huber, M., Hanebeck, U.: Analytic moment-based gaussian process filtering. In: Proceedings of the 26th ICML, pp. 225–232. ACM (2009)Google Scholar
  6. 6.
    Wang, J., Fleet, D., Hertzmann, A.: Gaussian process dynamical models for human motion. IEEE Transactions on PAMI 30, 283–298 (2008)CrossRefGoogle Scholar
  7. 7.
    Ko, J., Fox, D.: Learning gp-bayes filters via gaussian process latent variable models. Autonomous Robots 30, 3–23 (2011)CrossRefGoogle Scholar
  8. 8.
    Turner, R., Deisenroth, M., Rasmussen, C.: State-space inference and learning with gaussian processes. In: Proceedings of 13th AISTATS, vol. 9, pp. 868–875 (2010)Google Scholar
  9. 9.
    Nguyen-tuong, D., Peters, J.: Local gaussian process regression for real time online model learning and control. In: Advances in NIPS (2008)Google Scholar
  10. 10.
    Dempster, A.P., Laird, N.M., Rubin, D.B.: Maximum likelihood from incomplete data via the em algorithm. Journal of the Royal Statistical Society. Series B (Methodological), 1–38 (1977)Google Scholar
  11. 11.
    Ghahramani, Z., Hinton, G.: Parameter estimation for linear dynamical systems. Technical Report CRG-TR-96-2, University of Totronto (1996)Google Scholar
  12. 12.
    Hauskrecht, M., Valko, M., Batal, I., Clermont, G., Visweswaran, S., Cooper, G.: Conditional outlier detection for clinical alerting. In: AMIA Annual Symposium Proceedings, vol. 2010, pp. 286–290. AMIA (2010)Google Scholar
  13. 13.
    Hauskrecht, M., Batal, I., Valko, M., Visweswaran, S., Cooper, G.F., Clermont, G.: Outlier detection for patient monitoring and alerting. Journal of Biomedical Informatics 46(1), 47–55 (2013)CrossRefGoogle Scholar
  14. 14.
    Valko, M., Hauskrecht, M.: Feature importance analysis for patient management decisions. In: 13th International Congress on Medical Informatics, NIH Public Access, pp. 861–865 (2010)Google Scholar
  15. 15.
    Bibbona, E., Panfilo, G., Tavella, P.: The ornstein–uhlenbeck process as a model of a low pass filtered white noise. Metrologia 45, S117 (2008)MathSciNetCrossRefGoogle Scholar
  16. 16.
    Gibbs, M., MacKay, D.: Efficient implementation of gaussian processes (1997)Google Scholar
  17. 17.
    Kim, C.J.: Dynamic linear models with markov-switching. Journal of Econometrics 60, 1–22 (1994)MathSciNetzbMATHCrossRefGoogle Scholar
  18. 18.
    Hauskrecht, M., Fraser, H.: Modeling treatment of ischemic heart disease with partially observable markov decision processes. In: Proceedings of the AMIA Symposium, pp. 538–542. American Medical Informatics Association (1998)Google Scholar
  19. 19.
    Kveton, B., Hauskrecht, M.: Solving factored mdps with exponential-family transition models. In: 16th International Conference on Automated Planning and Scheduling, pp. 114–120 (2006)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  • Zitao Liu
    • 1
  • Milos Hauskrecht
    • 1
  1. 1.Department of Computer ScienceUniversity of PittsburghPittsburghUSA

Personalised recommendations