Advertisement

Neural Computing and Applications

, Volume 25, Issue 2, pp 269–281 | Cite as

A linear hybrid methodology for improving accuracy of time series forecasting

  • Ratnadip AdhikariEmail author
  • R. K. Agrawal
Original Article

Abstract

Modeling and forecasting of time series data are integral parts of many scientific and engineering applications. Increasing precision of the performed forecasts is highly desirable but a difficult task, facing a number of mathematical as well as decision-making challenges. This paper presents a novel approach for linearly combining multiple models in order to improve time series forecasting accuracy. Our approach is based on the assumption that each future observation of a time series is a linear combination of the arithmetic mean and median of the forecasts from all participated models together with a random noise. The proposed ensemble is constructed with five different forecasting models and is tested on six real-world time series. Obtained results demonstrate that the forecasting accuracies are significantly improved through our combination mechanism. A nonparametric statistical analysis is also carried out to show the superior forecasting performances of the proposed ensemble scheme over the individual models as well as a number of other forecast combination techniques.

Keywords

Time series Forecast combination Box-Jenkins models Artificial neural networks Elman networks Support vector machines 

Notes

Acknowledgments

The authors are thankful to the reviewers for their constructive suggestions which significantly facilitated the quality improvement of this paper. In addition, the first author likes to express his profound gratitude to the Council of Scientific and Industrial Research (CSIR), India, for the obtained financial support in performing this research work.

References

  1. 1.
    De Gooijer JG, Hyndman RJ (2006) 25 years of time series forecasting. J Forecast 22(3):443–473CrossRefGoogle Scholar
  2. 2.
    Zhang GP (2007) A neural network ensemble method with jittered training data for time series forecasting. Inform Sci 177(23):5329–5346CrossRefGoogle Scholar
  3. 3.
    Box GEP, Jenkins GM (1970) Time series analysis, forecasting and control, 3rd edn. Holden-Day, CaliforniazbMATHGoogle Scholar
  4. 4.
    Andrawis RR, Atiya AF, El-Shishiny H (2011) Forecast combinations of computational intelligence and linear models for the NN5 time series forecasting competition. Int J Forecast 27(3):672–688CrossRefGoogle Scholar
  5. 5.
    Lemke C, Gabrys B (2010) Meta-learning for time series forecasting and forecast combination. Neurocomputing 73:2006–2016CrossRefGoogle Scholar
  6. 6.
    Clemen RT (1989) Combining forecasts: a review and annotated bibliography. J Forecast 5(4):559–583CrossRefGoogle Scholar
  7. 7.
    Jose VRR, Winkler RL (2008) Simple robust averages of forecasts: some empirical results. Int J Forecast 24(1):163–169CrossRefGoogle Scholar
  8. 8.
    Armstrong JS (2001) Principles of forecasting: a handbook for researchers and practitioners. Academic Publishers, Norwell, MACrossRefGoogle Scholar
  9. 9.
    Stock JH, Watson MW (2004) Combination forecasts of output growth in a seven-country data set. J Forecast 23:405–430CrossRefGoogle Scholar
  10. 10.
    Larreche JC, Moinpour R (1983) Managerial judgment in marketing: the concept of expertise. J Market Res 20:110–121CrossRefGoogle Scholar
  11. 11.
    Agnew C (1985) Bayesian consensus forecasts of macroeconomic variables. J Forecast 4:363–376CrossRefGoogle Scholar
  12. 12.
    McNees SK (1992) The uses and abuses of ‘consensus’ forecasts. J Forecast 11:703–711CrossRefGoogle Scholar
  13. 13.
    Crane DB, Crotty JR (1967) A two-stage forecasting model: exponential smoothing and multiple regression. J Manage Sci 13:B501–B507CrossRefGoogle Scholar
  14. 14.
    Zarnowitz V (1967) An appraisal of short-term economic forecasts. National Bureau of Econ Res, New YorkGoogle Scholar
  15. 15.
    Reid DJ (1968) Combining three estimates of gross domestic product. J Economica 35:431–444CrossRefGoogle Scholar
  16. 16.
    Bates JM, Granger CWJ (1969) Combination of forecasts. J Oper Res Q 20:451–468CrossRefGoogle Scholar
  17. 17.
    Granger CWJ, Ramanathan R (1984) Improved methods of combining forecasts. J Forecast 3:197–204CrossRefGoogle Scholar
  18. 18.
    Granger CWJ, Newbold P (1986) Forecasting economic time series, 2nd edn. Academic Press, New YorkzbMATHGoogle Scholar
  19. 19.
    Aksu C, Gunter S (1992) An empirical analysis of the accuracy of SA, OLS, ERLS and NRLS combination forecasts. Int J Forecast 8:27–43CrossRefGoogle Scholar
  20. 20.
    Newbold P, Granger CWJ (1974) Experience with forecasting univariate time series and the combination of forecasts (with discussion). J R Stat Soc A 137(2):131–165CrossRefMathSciNetGoogle Scholar
  21. 21.
    Winkler RL, Makridakis S (1983) The combination of forecasts. J R Stat Soc A 146(2):150–157CrossRefGoogle Scholar
  22. 22.
    Pollock DSG (1993) Recursive estimation in econometrics. Comput Stat Data An 44:37–75CrossRefMathSciNetGoogle Scholar
  23. 23.
    Bunn D (1975) A Bayesian approach to the linear combination of forecasts. J Oper Res Q 26:325–329CrossRefzbMATHGoogle Scholar
  24. 24.
    Zhang GP (2003) Time series forecasting using a hybrid ARIMA and neural network model. Neurocomputing 50:159–175CrossRefzbMATHGoogle Scholar
  25. 25.
    Khashei M, Bijari M (2010) An artificial neural network (p, d, q) model for time series forecasting. J Expert Syst Appl 37(1):479–489CrossRefGoogle Scholar
  26. 26.
    Khashei M, Bijari M (2011) A novel hybridization of artificial neural networks and ARIMA models for time series forecasting. J Appl Soft Comput 11:2664–2675CrossRefGoogle Scholar
  27. 27.
    Adhikari R, Agrawal RK (2012) A novel weighted ensemble technique for time series forecasting. In: Proceedings of the 16th Pacific-Asia conference on knowledge discovery and data mining (PAKDD), Kuala Lumpur, Malaysia, pp 38–49Google Scholar
  28. 28.
    Winkler RL, Clemen RT (1992) Sensitivity of weights in combining forecasts. Oper Res 40(3):609–614CrossRefzbMATHGoogle Scholar
  29. 29.
    Tseng FM, Yu HC, Tzeng GH (2002) Combining neural network model with seasonal time series ARIMA model. J Technol Forecast Soc Change 69:71–87CrossRefGoogle Scholar
  30. 30.
    Vapnik V (1995) The nature of statistical learning theory. Springer, New YorkCrossRefzbMATHGoogle Scholar
  31. 31.
    Suykens JAK, Vandewalle J (1999) Least squares support vector machines classifiers. Neural Process Lett 9(3):293–300CrossRefMathSciNetGoogle Scholar
  32. 32.
    Hamzaçebi C, Akay D, Kutay F (2009) Comparison of direct and iterative artificial neural network forecast approaches in multi-periodic time series forecasting. J Expert Syst Appl 36(2):3839–3844CrossRefGoogle Scholar
  33. 33.
    Gheyas IA, Smith LS (2011) A novel neural network ensemble architecture for time series forecasting. Neurocomputing 74(18):3855–3864CrossRefGoogle Scholar
  34. 34.
    Lim CP, Goh WY (2005) The application of an ensemble of boosted Elman networks to time series prediction: a benchmark study. J Comput Intel 3(2):119–126Google Scholar
  35. 35.
    Chapelle O (2002) Support vector machines: introduction principles, adaptive tuning and prior knowledge. Ph.D. Thesis, University of Paris, FranceGoogle Scholar
  36. 36.
    Demuth H, Beale M, Hagan M (2010) Neural network toolbox user’s guide. The MathWorks, Natic, MAGoogle Scholar
  37. 37.
    Hyndman RJ (2011) Time series data library (TSDL), http://robjhyndman.com/TSDL/
  38. 38.
    Hollander M, Wolfe DA (1999) Nonparametric statistical methods. Wiley, Hoboken, NJzbMATHGoogle Scholar

Copyright information

© Springer-Verlag London 2013

Authors and Affiliations

  1. 1.School of Computer and Systems SciencesJawaharlal Nehru UniversityNew DelhiIndia

Personalised recommendations