Direct Multi-Step Estimation and Time Series Classification

  • Marcella CorduasEmail author


The AR metric represents a consolidated model-based approach for time series classification. The goodness of the final classification may of course be affected by the misspecification of the models describing the observed time series. This article investigates whether a direct multi-step estimation approach can shed some more light on time series comparison.


AR metric Time series classification Adaptive estimation Direct multi-step estimation 



This work was financially supported by PRIN project 2013-2015: “La previsione economica e finanziaria: il ruolo dell’informazione e la capacità di modellare il cambiamento”.


  1. 1.
    Bhansali, R.: Asymptotically efficient autoregressive model selection for multistep prediction. Ann. Inst. Stat. Math. 48, 577–602 (1996) CrossRefzbMATHMathSciNetGoogle Scholar
  2. 2.
    Bhansali, R.J.: Parameter estimation and model selection for multistep prediction of a time series: a review. In: Gosh, S. (ed.) Asymptotics, Nonparametrics, and Time Series, pp. 201–258. CRC Press, Boca Raton (1999) Google Scholar
  3. 3.
    Brockwell, A.P.J., Davies, R.A.: Time Series: Theory and Methods, 2nd edn. Springer, New York (1991) CrossRefGoogle Scholar
  4. 4.
    Chevillon, G.: Direct multi-step estimation and forecasting. J. Econ. Surv. 21, 746–785 (2007) CrossRefGoogle Scholar
  5. 5.
    Corduas, M.: La metrica autoregressiva tra modelli ARIMA: una procedura in linguaggio GAUSS. Quad. Stat. 2, 1–37 (2000) Google Scholar
  6. 6.
    Corduas, M.: Clustering streamflow time series for regional classification. J. Hydrol. 407, 73–80 (2011) CrossRefGoogle Scholar
  7. 7.
    Corduas, M., Piccolo, D.: Time series clustering and classification by the autoregressive metric. Comput. Stat. Data Anal. 52, 1860–1872 (2008) CrossRefzbMATHMathSciNetGoogle Scholar
  8. 8.
    Cox, D.R.: Prediction by exponentially weighted moving averages and related methods. J. R. Stat. Soc. B 23, 414–422 (1961) zbMATHGoogle Scholar
  9. 9.
    Marcellino, M., Stock, J.H., Watson, M.: A comparison of direct and iterated multistep AR methods for forecasting microeconomic time series. J. Econom. 135, 499–526 (2006) CrossRefMathSciNetGoogle Scholar
  10. 10.
    Piccolo, D.: Una topologia per la classe dei processi ARIMA. Statistica XLIV, 47–59 (1984) MathSciNetGoogle Scholar
  11. 11.
    Piccolo, D.: A distance measure for classifying ARIMA models. J. Time Ser. Anal. 11, 153–164 (1990) CrossRefzbMATHGoogle Scholar
  12. 12.
    Piccolo, D.: The autoregressive metric for comparing time series models. Statistica LXX, 459–480 (2010) Google Scholar
  13. 13.
    Proietti, T.: Direct and iterated multistep AR methods for difference stationary processes. Int. J. Forecast. 27, 266–280 (2011) CrossRefGoogle Scholar
  14. 14.
    Shibata, R.: Asymptotically efficient selection of the order of the model for estimating parameters of a linear process. Ann. Stat. 8, 147–164 (1980) CrossRefzbMATHGoogle Scholar
  15. 15.
    Tiao, G.C., Tsay, R.S.: Some advances in non-linear and adaptive modelling in time-series analysis. J. Forecast. 13, 109–131 (1994) CrossRefGoogle Scholar
  16. 16.
    Tiao, G.C., Xu, D.: Robustness of maximum likelihood estimates for multi-step predictions: the exponential smoothing case. Biometrika 80, 623–641 (1993) CrossRefzbMATHMathSciNetGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2014

Authors and Affiliations

  1. 1.Department of Political SciencesUniversity of Naples Federico IINaplesItaly

Personalised recommendations