This study compares the performance of statistical and Bayesian combination models with classical single time series models for short-term traffic forecasting. Combinations are based on fractionally integrated autoregressive time series models of travel speed with exogenous variables that consider speed's spatio-temporal evolution, and volume and weather conditions. Several statistical hypotheses on the effectiveness of combinations compared to the single models are also tested. Results show that, in the specific application, linear regression combination techniques may provide more accurate forecasts than Bayesian combination models. Moreover, combining models with different degrees of spatio-temporal complexity and exogeneities is most likely to be the best choice in terms of accuracy. Moreover, the risk of combining forecasts is lower than the risk of choosing a single model with increased spatio-temporal complexity.

References

1. 1)
  - 22. Van Lint, J.W.C., Van Hinsbergen, C.P.I.J.: ‘Short term traffic and travel time prediction models’, in Sadek, S., Chowdhury, R. (eds.): ‘Artificial intelligence applications to critical transportation issues’, (National Academies Press, Washington DC, 2012) Transportation Research Circular, Number E-C168, November.
2. 2)
  - E. Castillo , J.M. Menéndez , S. Sànchez-Cambronero . Predicting traffic flow using Bayesan networks. Transport. Res. B , 5 , 482 - 509
3. 3)
  - 2. Clemen, R.T.: ‘Combining forecasts: a review and annotated bibliography’, Int. J. Forecast., 1989, 5, pp. 559–583 (doi: 10.1016/0169-2070(89)90012-5).
4. 4)
  - 21. Karlaftis, M.G., Vlahogianni, E.I.: ‘Statistics versus neural networks in transportation research: differences, similarities and some insights’, Transp. Res. C, Emerg, Technol,, submitted, 2011, 19, (3), pp. 387–399 (doi: 10.1016/j.trc.2010.10.004).
5. 5)
  - 19. Zhang, Y.: ‘Hourly traffic forecasts using interacting multiple model (IMM) predictor’, IEEE Signal Process. Lett., 2011, 18, (10), pp. 607–610 (doi: 10.1109/LSP.2011.2165537).
6. 6)
  - 34. Vlahogianni, E.I., Yannis, G., Golias, J.C.: ‘Critical power two wheeler riding patterns at the emergence of an incident’, Accident Anal. Prevent., 2013, 58, pp. 340–345 (doi: 10.1016/j.aap.2012.12.026).
7. 7)
  - 32. Vlahogianni, E.I., Karlaftis, M.G., Golias, J.C., Halkias, B.: ‘Freeway operations, Spatio-temporal incident characteristics, and secondary-crash occurrence’, Transp. Res. Rec.: J. Transp. Res. Board, 2010, 2778, pp. 1–9 (doi: 10.3141/2178-01).
8. 8)
  - 17. Williams, B.M., Hoel, L.A.: ‘Modeling and forecasting vehicular traffic flow as a seasonal ARIMA process: theoretical basis and empirical results’, J. Transp. Eng., 2003, 129, (6), pp. 664–672 (doi: 10.1061/(ASCE)0733-947X(2003)129:6(664)).
9. 9)
  - 1. Bates, J.M., Granger, C.W.J.: ‘The combination of forecasts’. Oper. Res. Q., 1969, 20, pp. 451–468 (doi: 10.1057/jors.1969.103).
10. 10)
  - 18. Huang, H., Lee, T.H.: ‘To combine forecasts or to combine information?’, Econ. Rev., 2010, 29, (5–6), pp. 534–570 (doi: 10.1080/07474938.2010.481553).
11. 11)
  - 9. Tan, M.C., Wong, S.C., Xu, J.M., Guan, Z.R., Zhang, P.: ‘An aggregation approach to short-term traffic flow prediction’, IEEE Trans. Intell. Transp. Syst., 2009, 10, (1), pp. 60–69 (doi: 10.1109/TITS.2008.2011693).
12. 12)
  - W. Zheng , D. Lee , Q. Shi . Short-term freeway traffic flow prediction: Bayesian combined neural network approach. J. Transp. Eng. , 2 , 114 - 121
13. 13)
  - 3. Smith, J., Wallis, K.F.: ‘A simple explanation of the forecast combination puzzle’, Oxford Bull. Econ. Stat., 2009, 71, pp. 331–355 (doi: 10.1111/j.1468-0084.2008.00541.x).
14. 14)
  - H. Akaike . A new look at statistical model identification. IEEE Trans. Autom. Control , 716 - 723
15. 15)
  - 6. Yang, Y.: ‘Combining forecasting procedures: some theoretical results’, Econ. Theory, 2004, 20, pp. 176–222 (doi: 10.1017/S0266466604201086).
16. 16)
  - 24. Vlahogianni, E.I., Karlaftis, M.G.: ‘Testing and comparing neural network and statistical approaches for predicting transportation time series’, Transp. Res. Rec., J, Transp. Res. Board, 2013, 2399, (1), pp. 9–22 (doi: 10.3141/2399-02).
17. 17)
  - 5. Rapach, D.E., Strauss, J.K.: ‘Bagging or combining (or both)?’, an analysis based on forecasting U.S. employment growth’, Econ. Rev., 2010, 29, (5–6), pp. 511–533 (doi: 10.1080/07474938.2010.481550).
18. 18)
  - 13. Stathopoulos, A., Dimitriou, L., Tsekeris, T.: ‘Fuzzy modeling approach for combined forecasting of urban traffic flow’, Comput. Aided Civil Infrastruct. Eng., 2008, 23, (7), pp. 521–535 (doi: 10.1111/j.1467-8667.2008.00558.x).
19. 19)
  - 10. Vlahogianni, E.I., Golias, J.C., Karlaftis, M.G.: ‘Short-term traffic forecasting: overview of objectives and methods’, Transp. Rev., 2004, 24, (5), pp. 533–557 (doi: 10.1080/0144164042000195072).
20. 20)
  - 25. Washington, S.P., Karlaftis, M.G., Mannering, F.L.: ‘Statistical and econometric methods for transportation data analysis’ (CRC press, 2010).
21. 21)
  - G. Schwarz . Estimating the dimension of a model. Ann. Stat. , 2 , 461 - 464
22. 22)
  - S. Sun , C. Zhang , G. Yu . A Bayesian network approach to traffic flow forecasting. IEEE Trans. Intell. Transp. Syst. , 1 , 124 - 132
23. 23)
  - 23. Vlahogianni, E.I., Karlaftis, M.G., Golias, J.C.: ‘Short-term traffic forecasting: where we are and where we're going’, Transp. Res. C, Emerg. Technol., 2014, 43, (1), pp. 3–19 (doi: 10.1016/j.trc.2014.01.005).
24. 24)
  - 9. Song, H., Witt, S.F., Wu, D.C., Wong, K.K.F.: ‘Tourism forecasting: to combine or not to combine?’, Tour. Manage., 2006, 28, pp. 1068–1078.
25. 25)
  - 29. Friedman, N., Geiger, D., Goldszmidt, M.: ‘Bayesian network classifiers’, Mach. Learn., 1997, 29, pp. 131–163 (doi: 10.1023/A:1007465528199).
26. 26)
  - 33. Vlahogianni, E.I., Golias, J.C.: ‘Bayesian modeling of the microscopic traffic characteristics of overtaking in two-lane highways’, Transp. Res. F. Traffic Psychol. Behav., 2012, 15, (3), pp. 348–357 (doi: 10.1016/j.trf.2012.02.002).
27. 27)
  - 15. Djuric, N., Radosavljevic, V., Coric, V., Vucetic, S.: ‘Travel speed forecasting by means of continuous conditional random fields’, Transp. Res. Rec., J. Transp. Res. Board, 2011, 2263, (1), pp. 131–139 (doi: 10.3141/2263-15).
28. 28)
  - 30. Vlahogianni, E.I., Webber, Jr.C.L., Geroliminis, N., Skabardonis, A.: ‘Statistical characteristics of transitional queue conditions in signalized arterials’, Transp. Res. Part C, Emerg. Technol., 2007, 15, (6), pp. 345–404 (doi: 10.1016/j.trc.2007.07.003).
29. 29)
  - 35. Diebold, F.X., Mariano, R.S.: ‘Comparing predictive accuracy’, J. Bus. Econ. Stat., 1995, 13, pp. 253–263.
30. 30)
  - 7. Wei, X., Yang, Y.: ‘Robust forecast combinations’, J. Econ., 2012, 166, pp. 224–236 (doi: 10.1016/j.jeconom.2011.09.035).
31. 31)
  - 10. Vlahogianni, E.I.: ‘Enhancing predictions in signalized arterials with information on short-term traffic flow dynamics’, J. Intell. Transp. Syst., 2009, 13, (2), pp. 73–84 (doi: 10.1080/15472450902858384).
32. 32)
  - 8. Hibon, M., Evgeniou, T.: ‘To combine or not to combine: selecting among forecasts and their combinations’, Int. J. Forecast., 2005, 21, pp. 15–24 (doi: 10.1016/j.ijforecast.2004.05.002).
33. 33)
  - R. Chrobok , O. Kaumann , J. Wahle , M. Schreckenberg . Different methods of traffic forecast based on real data. Eur. J. Oper. Res. , 3 , 558 - 568
34. 34)
  - 4. Diebold, F.X., Pauly, P.: ‘The use of prior information in forecast combination’, Int. J. Forecast., 1990, 6, pp. 503–508 (doi: 10.1016/0169-2070(90)90028-A).
35. 35)
  - 28. Karlaftis, M.G., Vlahogianni, E.I.: ‘Memory properties and fractional integration in transportation time-series’, Transp. Res. C. Emerg. Technol., 2009, 17, (4), pp. 444–453 (doi: 10.1016/j.trc.2009.03.001).

Improving short-term traffic forecasts: to combine models or not to combine?

References

Related content