Skip to main content
SpringerLink
Log in

Bayesian forecasting of multivariate time series: scalability, structure uncertainty and decisions

  • Invited Article: Second Akaike Memorial Lecture
  • Published:
Annals of the Institute of Statistical Mathematics Aims and scope Submit manuscript

Abstract

I discuss recent research advances in Bayesian state-space modeling of multivariate time series. A main focus is on the “decouple/recouple” concept that enables application of state-space models to increasingly large-scale data, applying to continuous or discrete time series outcomes. Applied motivations come from areas such as financial and commercial forecasting and dynamic network studies. Explicit forecasting and decision goals are often paramount and should factor into model assessment and comparison, a perspective that is highlighted. The Akaike Memorial Lecture is a context to reflect on the contributions of Hirotugu Akaike and to promote new areas of research. In this spirit, this paper aims to promote new research on foundations of statistics and decision analysis, as well as on further modeling, algorithmic and computational innovation in dynamic models for increasingly complex and challenging problems in multivariate time series analysis and forecasting.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14

Similar content being viewed by others

References

  • Aguilar, O., West, M. (2000). Bayesian dynamic factor models and portfolio allocation. Journal of Business and Economic Statistics, 18, 338–357.

    Google Scholar 

  • Aguilar, O., Prado, R., Huerta, G., West, M. (1999). Bayesian inference on latent structure in time series. In J. M. Bernardo, J. O. Berger, A. P. Dawid and A. F. M. Smith (Eds.), Bayesian statistics, Vol. 6, pp. 3–26. Oxford: Oxford University Press, (with discussion).

  • Ahelegbey, D. F., Billio, M., Casarin, R. (2016a). Bayesian graphical models for structural vector autoregressive processes. Journal of Applied Econometrics, 31, 357–386.

    MathSciNet  Google Scholar 

  • Ahelegbey, D. F., Billio, M., Casarin, R. (2016b). Sparse graphical multivariate autoregression: A Bayesian approach. Annals of Economics and Statistics, 123(124), 1–30.

    Google Scholar 

  • Akaike, H. (1974). A new look at the statistical model identification. IEEE Transactions on Automatic Control, 19, 716–723.

    MathSciNet  MATH  Google Scholar 

  • Akaike, H. (1978). On the likelihood of a time series model. Journal of the Royal Statistical Society (Series D: The Statistician), 27, 217–235.

  • Akaike, H. (1979). A Bayesian extension of the minimum AIC procedure of autoregressive model fitting. Biometrika, 66, 237–242.

    MathSciNet  MATH  Google Scholar 

  • Akaike, H. (1981). Likelihood of a model and information criteria. Journal of Econometrics, 16, 3–14.

    MATH  Google Scholar 

  • Aktekin, T., Polson, N. G., Soyer, R. (2018). Sequential Bayesian analysis of multivariate count data. Bayesian Analysis, 13, 385–409.

    MathSciNet  MATH  Google Scholar 

  • Anacleto, O., Queen, C. M., Albers, C. J. (2013). Multivariate forecasting of road traffic flows in the presence of heteroscedasticity and measurement errors. Journal of the Royal Statistical Society (Series C: Applied Statistics), 62, 251–270.

    Google Scholar 

  • Berry, L. R., West, M. (2019). Bayesian forecasting of many count-valued time series. Journal of Business and Economic Statistics, https://doi.org/10.1080/07350015.2019.1604372.

  • Berry, L. R., Helman, P., West, M. (2019). Probabilistic forecasting of heterogeneous consumer transaction-sales time series. International Journal of Forecasting, https://doi.org/10.1016/j.ijforecast.2019.07.007.

  • Bianchi, D., Billio, M., Casarin, R., Guidolin, M. (2019). Modeling systemic risk with Markov switching graphical SUR models. Journal of Econometrics, 210, 58–74.

    MathSciNet  MATH  Google Scholar 

  • Bodkin, R. G., Klein, L. R., Marwah, K. (1991). A history of macroeconometric model-building. Aldershot: Edward Elgar.

    Google Scholar 

  • Cargnoni, C., Müller, P., West, M. (1997). Bayesian forecasting of multinomial time series through conditionally Gaussian dynamic models. Journal of the American Statistical Association, 92, 640–647.

    MATH  Google Scholar 

  • Carvalho, C. M., West, M. (2007a). Dynamic matrix-variate graphical models. Bayesian Analysis, 2, 69–98.

    MathSciNet  MATH  Google Scholar 

  • Carvalho, C. M., West, M. (2007b). Dynamic matrix-variate graphical models-A synopsis. In J. M. Bernardo, M. J. Bayarri, J. O. Berger, A. P. Dawid, D. Heckerman, A. F. M. Smith and M. West (Eds.), Bayesian statistics, Vol. 8, pp. 585–590. Oxford: Oxford University Press.

  • Carvalho, C. M., Massam, H., West, M. (2007). Simulation of hyper-inverse Wishart distributions in graphical models. Biometrika, 94, 647–659.

    MathSciNet  MATH  Google Scholar 

  • Carvalho, C. M., Lopes, H. F., Aguilar, O. (2011). Dynamic stock selection strategies: A structured factor model framework (with discussion). In J. M. Bernardo, M. J. Bayarri, J. O. Berger, A. P. Dawid, D. Heckerman, A. F. M. Smith and M. West (Eds.), Bayesian statistics, Vol. 9, pp. 69–90. Oxford: Oxford University Press.

  • Chen, C. W. S., Lee, S. (2017). Bayesian causality test for integer-valued time series models with applications to climate and crime data. Journal of the Royal of Statistical Society (Series C: Applied Statistics), 66, 797–814.

    MathSciNet  Google Scholar 

  • Chen, X., Irie, K., Banks, D., Haslinger, R., Thomas, J., West, M. (2018). Scalable Bayesian modeling, monitoring and analysis of dynamic network flow data. Journal of the American Statistical Association, 113, 519–533.

    MathSciNet  Google Scholar 

  • Chen, X., Banks, D., West, M. (2019). Bayesian dynamic modeling and monitoring of network flows. Network Science, 7, 292–318.

    Google Scholar 

  • Clyde, M., Iversen, E. S. (2013). Bayesian model averaging in the M-open framework. In P. Damien, P. Dellaportes, N. G. Polson and D. A. Stephens (Eds.), Bayesian theory and applications, pp. 484–498. Oxford: Oxford University Press.

  • Congdon, P. (2000). A Bayesian approach to prediction using the gravity model, with an application to patient flow modeling. Geographical Analysis, 32, 205–224.

    Google Scholar 

  • Costa, L., Smith, J. Q., Nichols, T., Cussens, J., Duff, E. P., Makin, T. R. (2015). Searching multiregression dynamic models of resting-state fMRI networks using integer programming. Bayesian Analysis, 10, 441–478.

    MathSciNet  MATH  Google Scholar 

  • Del Negro, M., Otrok, C. M. (2008) Dynamic factor models with time-varying parameters: Measuring changes in international business cycles. Staff Report 326, New York Federal Reserve.

  • Ferreira, M. A. R., Gamerman, D., Migon, H. S. (1997). Bayesian dynamic hierarchical models: Covariance matrices estimation and nonnormality. Brazilian Journal of Probability and Statistics, 11, 67–79.

    MATH  Google Scholar 

  • Ferreira, M. A. R., West, M., Lee, H., Higdon, D. M. (2006). Multiscale and hidden resolution time series models. Bayesian Analysis, 2, 294–314.

    Google Scholar 

  • Gamerman, D., Migon, H. S. (1993). Dynamic hierarchical models. Journal of the Royal Statistical Society (Series B: Methodological), 55, 629–642.

    MathSciNet  MATH  Google Scholar 

  • Glynn, C., Tokdar, S. T., Banks, D. L., Howard, B. (2019). Bayesian analysis of dynamic linear topic models. Bayesian Analysis, 14, 53–80.

    MathSciNet  MATH  Google Scholar 

  • Gruber, L. F. (2019). github.com/lutzgruber/gpuSGDLM.

  • Gruber, L. F., West, M. (2016). GPU-accelerated Bayesian learning in simultaneous graphical dynamic linear models. Bayesian Analysis, 11, 125–149.

    MathSciNet  MATH  Google Scholar 

  • Gruber, L. F., West, M. (2017). Bayesian forecasting and scalable multivariate volatility analysis using simultaneous graphical dynamic linear models. Econometrics and Statistics, 3, 3–22.

    MathSciNet  Google Scholar 

  • Hans, C., Dobra, A., West, M. (2007a). Shotgun stochastic search in regression with many predictors. Journal of the American Statistical Association, 102, 507–516.

    MathSciNet  MATH  Google Scholar 

  • Hans, C., Wang, Q., Dobra, A., West, M. (2007b). SSS: High-dimensional Bayesian regression model search. Bulletin of the International Society for Bayesian Analysis, 24, 8–9.

    Google Scholar 

  • Hazelton, M. L. (2015). Network tomography for integer-valued traffic. Annals of Applied Statistics, 9, 474–506.

    MathSciNet  MATH  Google Scholar 

  • Irie, K., West, M. (2019). Bayesian emulation for multi-step optimization in decision problems. Bayesian Analysis, 14, 137–160.

    MathSciNet  MATH  Google Scholar 

  • Jandarov, R., Haran, M., Bjornstad, O. N., Grenfell, B. T. (2014). Emulating a gravity model to infer the spatiotemporal dynamics of an infectious disease. Journal of the Royal Statistical Society (Series C: Applied Statistics), 63, 423–444.

    MathSciNet  Google Scholar 

  • Jones, B., West, M. (2005). Covariance decomposition in undirected Gaussian graphical models. Biometrika, 92, 779–786.

    MathSciNet  MATH  Google Scholar 

  • Jones, B., Dobra, A., Carvalho, C. M., Hans, C., Carter, C., West, M. (2005). Experiments in stochastic computation for high-dimensional graphical models. Statistical Science, 20, 388–400.

    MathSciNet  MATH  Google Scholar 

  • Kastner, G., Frühwirth-Schnatter, S., Lopes, H. F. (2017). Efficient Bayesian inference for multivariate factor stochastic volatility models. Journal of Computational and Graphical Statistics, 26, 905–917.

    MathSciNet  Google Scholar 

  • Kitagawa, G., Gersch, W. (1996). Smoothness priors analysis of time series. Lecture Notes in Statistics New York: Springer.

    MATH  Google Scholar 

  • Kliesen, K. L., Smith, D. C. (2010). Measuring financial market stress: The St. Louis Fed’s financial stress index (STLFSI). In Federal Reserve Bank of St Louis National Economic Trends.

  • Konishi, S., Kitagawa, G. (2007). Information criteria and statistical modeling. New York: Springer.

    MATH  Google Scholar 

  • Koop, G., Korobilis, D. (2010). Bayesian multivariate time series methods for empirical macroeconomics. Foundations and Trends in Econometrics, 3, 267–358.

    MATH  Google Scholar 

  • Koop, G., Korobilis, D. (2013). Large time-varying parameter VARs. Journal of Econometrics, 177, 185–198.

    MathSciNet  MATH  Google Scholar 

  • Lavine, I., Lindon, M., West, M. (2019). Adaptive variable selection for sequential prediction in multivariate dynamic models. Technical Report, Department of Statistical Science, Duke University, Durham, North Carolina, ArXiv:1906.06580.

  • Liu, F., West, M. (2009). A dynamic modelling strategy for Bayesian computer model emulation. Bayesian Analysis, 4, 393–412.

    MathSciNet  MATH  Google Scholar 

  • Lopes, H. F., Carvalho, C. M. (2007). Factor stochastic volatility with time varying loadings and Markov switching regimes. Journal of Statistical Planning and Inference, 137, 3082–3091.

    MathSciNet  MATH  Google Scholar 

  • Lopes, H. F., McCulloch, R. E., Tsay, R. S. (2018) Parsimony inducing priors for large scale state-space models. Technical Report 2018-08, Booth School of Business, University of Chicago, Chicago, Illinois.

  • McAlinn, K., West, M. (2019). Dynamic Bayesian predictive synthesis in time series forecasting. Journal of Econometrics, 210, 155–169.

    MathSciNet  MATH  Google Scholar 

  • McAlinn, K., Aastveit, K. A., West, M. (2018). Bayesian predictive synthesis-discussion of: Using stacking to average Bayesian predictive distributions. Bayesian Analysis, 13, 971–973.

    Google Scholar 

  • McAlinn, K., Aastveit, K. A., Nakajima, J., West, M. (2019). Multivariate Bayesian predictive synthesis in macroeconomic forecasting. Journal of the American Statistical Association, https://doi.org/10.1080/01621459.2019.1660171.

  • Nakajima, J., West, M. (2013a). Bayesian analysis of latent threshold dynamic models. Journal of Business and Economic Statistics, 31, 151–164.

    MathSciNet  Google Scholar 

  • Nakajima, J., West, M. (2013b). Bayesian dynamic factor models: Latent threshold approach. Journal of Financial Econometrics, 11, 116–153.

    Google Scholar 

  • Nakajima, J., West, M. (2015). Dynamic network signal processing using latent threshold models. Digital Signal Processing, 47, 6–15.

    MathSciNet  Google Scholar 

  • Nakajima, J., West, M. (2017). Dynamics and sparsity in latent threshold factor models: A study in multivariate EEG signal processing. Brazilian Journal of Probability and Statistics, 31, 701–731.

    MathSciNet  MATH  Google Scholar 

  • Parzen, E., Tanabe, K., Kitagawa, G. (Eds.) (1998). Selected papers of Hirotugu Akaike. New York: Springer.

    MATH  Google Scholar 

  • Prado, R., West, M. (2010). Time series: Modeling, computation and inference. Boca Raton: Chapman & Hall.

  • Prado, R., Molina, F. J., Huerta, G. (2006). Multivariate time series modeling and classification via hierarchical VAR mixtures. Computational Statistics and Data Analysis, 51, 1445–1462.

    MathSciNet  MATH  Google Scholar 

  • Primiceri, G. E. (2005). Time varying structural vector autoregressions and monetary policy. Review of Economic Studies, 72, 821–852.

    MathSciNet  MATH  Google Scholar 

  • Queen, C. M. (1994). Using the multiregression dynamic model to forecast brand sales in a competitive product market. Journal of the Royal Statistical Society (Series D: The Statistician), 43, 87–98.

  • Queen, C. M., Smith, J. Q. (1993). Multiregression dynamic models. Journal of the Royal Statistical Society (Series B: Methodological), 55, 849–870.

  • Queen, C. M., Wright, B. J., Albers, C. J. (2008). Forecast covariances in the linear multiregression dynamic model. Journal of Forecasting, 27, 175–191.

    MathSciNet  Google Scholar 

  • Raftery, A. E., Kárný, M., Ettler, P. (2010). Online prediction under model uncertainty via dynamic model averaging: Application to a cold rolling mill. Technometrics, 52, 52–66.

    MathSciNet  Google Scholar 

  • Reeson, C., Carvalho, C. M., West, M. (2009). Dynamic graphical models and portfolio allocations for structured mutual funds. Discussion Paper 2009–27, Duke University, Duke University, Durham, North Carolina.

  • Scott, J. G., Carvalho, C. M. (2008). Feature-inclusion stochastic search for Gaussian graphical models. Journal of Computational and Graphical Statistics, 17, 790–808.

    MathSciNet  Google Scholar 

  • Sen, A., Smith, T. (1995). Gravity models of spatial interaction behavior. New York: Springer.

    Google Scholar 

  • Shirota, S., Omori, Y., Lopes, H. F., Piao, H. (2017). Cholesky realized stochastic volatility model. Econometrics and Statistics, 3, 34–59.

    Google Scholar 

  • Smith, M., Kohn, R. (2002). Parsimonious covariance matrix estimation for longitudinal data. Journal of the American Statistical Association, 97, 1141–1153.

    MathSciNet  MATH  Google Scholar 

  • Tank, A., Foti, N. J., Fox, E. B. (2015). Bayesian structure learning for stationary time series. In M. Meila., T. Heskes (Eds.), Proceedings of the 31st conference on uncertainty in artificial intelligence (UAI) (pp. 872–881), http://www.auai.org/uai2015.

  • Tebaldi, C., West, M. (1998). Bayesian inference on network traffic using link count data. Journal of the American Statistical Association, 93, 557–573.

    MathSciNet  MATH  Google Scholar 

  • Tebaldi, C., West, M., Karr, A. F. (2002). Statistical analyses of freeway traffic flows. Journal of Forecasting, 21, 39–68.

    Google Scholar 

  • Terui, N., Ban, M. (2014). Multivariate time series model with hierarchical structure for over-dispersed discrete outcomes. Journal of Forecasting, 33, 379–390.

    MathSciNet  MATH  Google Scholar 

  • Triantafyllopoulos, K. (2009). Inference of dynamic generalized linear models: On-line computation and appraisal. International Statistical Review, 77, 430–450.

    Google Scholar 

  • Walker, S. G., Gutiérrez-Peña, E., Muliere, P. (2001). A decision theoretic approach to model averaging. Journal of the Royal Statistical Society (Series D: The Statistician), 50, 31–39.

  • Wang, H. (2010). Sparse seemingly unrelated regression modelling: Applications in econometrics and finance. Computational Statistics and Data Analysis, 54, 2866–2877.

    MathSciNet  MATH  Google Scholar 

  • Wang, H. (2015). Scaling it up: Stochastic search structure learning in graphical models. Bayesian Analysis, 10, 351–377.

    MathSciNet  MATH  Google Scholar 

  • Wang, H., West, M. (2009). Bayesian analysis of matrix normal graphical models. Biometrika, 96, 821–834.

    MathSciNet  MATH  Google Scholar 

  • Wang, H., Reeson, C., Carvalho, C. M. (2011). Dynamic financial index models: Modeling conditional dependencies via graphs. Bayesian Analysis, 6, 639–664.

    MathSciNet  MATH  Google Scholar 

  • West, M. (1994). Statistical inference for gravity models in transportation flow forecasting. Discussion Paper 1994–40, Duke University, Durham, North Carolina, and Technical Report #60, National Institute of Statistical Sciences, Research Triangle Park, North Carolina.

  • West, M. (2013). Bayesian dynamic modelling. In P. Damien, P. Dellaportes, N. G. Polson and D. A. Stephens (Eds.), Bayesian theory and applications, pp. 145–166. Oxford: Oxford University Press.

  • West, M., Harrison, P. J. (1986). Monitoring and adaptation in Bayesian forecasting models. Journal of the American Statistical Association, 81, 741–750.

    MATH  Google Scholar 

  • West, M., Harrison, P. J. (1989a). Bayesian forecasting and dynamic models, 1st ed. New York: Springer.

  • West, M., Harrison, P. J. (1989b). Subjective intervention in formal models. Journal of Forecasting, 8, 33–53.

    Google Scholar 

  • West, M., Harrison, P. J. (1997). Bayesian forecasting and dynamic models, 2nd ed. New York: Springer.

  • West, M., Harrison, P. J., Migon, H. S. (1985). Dynamic generalised linear models and Bayesian forecasting. Journal of the American Statistical Association, 80, 73–97, (with discussion).

    MathSciNet  MATH  Google Scholar 

  • Xie, M. (2012). Discount-weighted Bayesian model averaging for portfolio decisions in matrix variate dynamic linear models. Undergraduate Thesis, Department of Statistical Science, Duke University, Durham, North Carolina.

  • Yao, Y., Vehtari, A., Simpson, D., Gelman, A. (2018). Using stacking to average Bayesian predictive distributions. Bayesian Analysis, 13, 917–1007, (with discussion).

    MathSciNet  MATH  Google Scholar 

  • Yelland, P. M. (2009). Bayesian forecasting for low-count time series using state-space models: An empirical evaluation for inventory management. International Journal of Production Economics, 118, 95–103.

    Google Scholar 

  • Zhao, Z. Y., Xie, M., West, M. (2016). Dynamic dependence networks: Financial time series forecasting and portfolio decisions. Applied Stochastic Models in Business and Industry, 32, 311–339, (with discussion).

    MathSciNet  MATH  Google Scholar 

  • Zhou, X., Nakajima, J., West, M. (2014). Bayesian forecasting and portfolio decisions using dynamic dependent sparse factor models. International Journal of Forecasting, 30, 963–980.

    Google Scholar 

Download references

Acknowledgements

I was honored to be invited by the Institute of Statistical Mathematics and the Japan Statistical Society to present the 2018 Akaike Memorial Lecture. This paper concerns research featured in that address, presented at the Annual Conference of the Japanese Federation of Statistical Science Associations, Tokyo, Japan, on September 10, 2018. I acknowledge the Akaike Memorial Lecture Award committee and the meeting conveners, and constructive comments of invited discussants Chris Glynn and Jouchi Nakajima. Additional thanks go to the past students and collaborators on topics touched on in this paper, many noted as co-authors in the reference list. Particular thanks are due to Lindsay Berry, Xi Chen and Lutz Gruber on some recent examples and suggestions.

Author information

Authors and Affiliations

  1. Department of Statistical Science, Duke University, Durham, NC, 27708-0251, USA

    Mike West

Authors
  1. Mike West
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Mike West.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

The Related Articles are https://doi.org/10.1007/s10463-019-00742-2; https://doi.org/10.1007/s10463-019-00743-1; https://doi.org/10.1007/s10463-019-00744-0.

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

West, M. Bayesian forecasting of multivariate time series: scalability, structure uncertainty and decisions. Ann Inst Stat Math 72, 1–31 (2020). https://doi.org/10.1007/s10463-019-00741-3

Download citation

  • Received:

  • Revised:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10463-019-00741-3

Keywords

Search

Navigation