ABSTRACT
Sensored traffic data in modern cities have been collected and applied for various purposes in the domain of intelligent transportation systems (ITS). However, analyzing these traffic data often lacks in priori knowledge due to the dynamics of transportation systems, making it hard to cope with diverse scenarios with specific models. In view of the limitations of traditional approaches, in this paper, we propose the Stepwise Heterogeneous Ensemble (SHE) for citywide traffic analysis based on stacked generalization. We first prove SHE's effectiveness using error-ambiguity decomposition technique. Secondly we analyze the optimal linear combination of SHE and present the stepwise iterating strategy. We also demonstrate its validity based on Kullback-Leibler divergence analysis. Thirdly we integrate six classical approaches into SHE framework, including linear least squares regression (LLSR), autoregressive moving average (ARMA), historical mean (HM), artificial neural network (ANN), radical basis function neural network (RBFNN), support vector machine (SVM). We further compare SHE's performance with other four linear combination models, namely equal weights method (EW), optimal weights method (OW), minimum error method (ME) and minimum variance method (MV). A series of experiments are conducted with a real city traffic dataset in Beijing city. The results show that the proposed SHE method behaves more robust and precise than other six single methods. Moreover, this method also outperforms other four different combination strategies both in variance and bias. In addition, the SHE method provides an open-ending framework for citywide traffic analysis, which means any new promising models can be easily incorporated into it in the future.
- Fusco, G., Colombaroni, C., Isaenko, N. 2016. Short-term speed predictions exploiting big data on large urban road networks. TRANSPORT RES C-EMER, 73, 183--201.Google ScholarCross Ref
- Zheng, F., Van Zuylen, H. 2013. Urban link travel time estimation based on sparse probe vehicle data. TRANSPORT RES C-EMER, 31, 145--157.Google ScholarCross Ref
- Rajabzadeh, Y., Rezaie, A. H., Amindavar, H. 2017. Short-term traffic flow prediction using time-varying Vasicek model. TRANSPORT RES C-EMER, 74, 168--181.Google ScholarCross Ref
- Huang, Y., Zhao, L., Van Woensel, T., Gross, J. P. 2017. Time-dependent vehicle routing problem with path flexibility. TRANSPORT RES B-METH, 95, 169--195.Google ScholarCross Ref
- Liu, X., Liu, K., Li, M., Lu, F. 2017. A ST-CRF Map-Matching Method for Low-Frequency Floating Car Data. IEEE T INTELL TRANSP, 18(5), 1241--1254. Google ScholarDigital Library
- Ahmed, K., Abu-Lebdeh, G., Al-Omari, B. 2012. Estimation of delay induced by downstream operations at signalized intersections over extended control time. J TRANSP ENG-ASCE, 139(1), 8--19.Google ScholarCross Ref
- Zheng, Y. 2015. Trajectory data mining: an overview. ACM T INTEL SYST TEC, 6(3), 29 Google ScholarDigital Library
- Ban, X., Herring, R., Hao, P., Bayen, A. 2009. Delay pattern estimation for signalized intersections using sampled travel times. TRANSPORT RES REC, 2130, 109--119.Google ScholarCross Ref
- Zhang, Y., Liu, Y. 2011. Analysis of peak and non-peak traffic forecasts using combined models. J ADV TRANSPORT, 45(1), 21--37.Google ScholarCross Ref
- Ambühl, L., Menendez, M. 2016. Data fusion algorithm for macroscopic fundamental diagram estimation. TRANSPORT RES C-EMER, 71, 184--197..Google ScholarCross Ref
- Long, J., Gao, Z., Zhao, X., Lian, A., Orenstein, P. 2011. Urban traffic jam simulation based on the cell transmission model. NETW SPAT ECON, 11(1), 43--64.Google ScholarCross Ref
- Hibon, M., Evgeniou, T. 2005. To combine or not to combine: selecting among forecasts and their combinations. INT J FORECASTING, 21(1), 15--24.Google ScholarCross Ref
- Wolpert, D. H. 1992. Stacked generalization. NEURAL NETWORKS, 5(2), 241--259. Google ScholarDigital Library
- Krogh, A., Vedelsby, J. 1995. Neural network ensembles, cross validation, and active learning. NIPS, 7, 231--238. Google ScholarDigital Library
- Galas, D. J., Dewey, T. G., Kunert-Graf, J., Sakhanenko, N. A. 2017. Expansion of the Kullback-Leibler Divergence, and a new class of information metrics. arXiv preprint arXiv:1702.00033.Google Scholar
- João, M.M, Carlos, S., Alípio, M.J. and Jorge, F. 2012. Ensemble approaches for regression: A survey. ACM Comput. Surv. 45, 1. Google ScholarDigital Library
- Zhou, Z.H. 2012. Ensemble Methods: Foundations and Algorithms, Boca Raton, FL: Chapman & Hall/CRC, 12--31. Google ScholarCross Ref
- Zhang, M. L., Zhou, Z. H. 2013. Exploiting unlabeled data to enhance ensemble diversity. DATA MIN KNOWL DISC, 26(1), 98--129. Google ScholarDigital Library
- Krawczyk, B., Minku, L. L., Gama, J., Stefanowski, J., Woźniak, M. 2017. Ensemble learning for data stream analysis: a survey. INFORM FUSION, 37, 132--156. Google ScholarDigital Library
- Branco, P., Torgo, L., Ribeiro, R. P. 2016. A survey of predictive modeling on imbalanced domains. ACM Comput. Surv, 49(2), 31. Google ScholarDigital Library
- Chen, Y., Wong, M. L., Li, H. 2014. Applying Ant Colony Optimization to configuring stacking ensembles for data mining. EXPERT SYST APPL, 41(6), 2688--2702. Google ScholarDigital Library
- King, M. A., Abrahams, A. S., Ragsdale, C. T. 2014. Ensemble methods for advanced skier days prediction. EXPERT SYST APPL, 41(4), 1176--1188. Google ScholarDigital Library
- Andreas Töscher, Michael Jahrer, Robert M. Bell, The BigChaos Solution to the Netflix Grand Prize, Report from the Netflix Prize Winners, 2009.Google Scholar
- Marc Claesen, Frank De Smet, Johan A.K. Suykens, Bart De Moor. 2014. EnsembleSVM: A Library for Ensemble Learning Using Support Vector Machines. J MACH LEARN RES. 15, 141--145. Google ScholarDigital Library
- Heitor Murilo Gomes, Jean Paul Barddal, Fabrício Enembreck, Albert Bifet. A Survey on Ensemble Learning for Data Stream Classification. ACM Comput. Surv. 50(2), 23. Google ScholarDigital Library
- Bartosz Krawczyk, Leandro L. Minkub, João Gamac, Jerzy Stefanowskid, Michał Woźniake. 2017. Ensemble learning for data stream analysis: A survey. INFORM FUSION, 37, 132--156. Google ScholarDigital Library
- Robert E. Schapire. 1990. The strength of weak learnability. MACH LEARN. 5(2), 197--227. Google ScholarDigital Library
- Nascimento D.S.C., Coelho A.L.V. 2009. Ensembling Heterogeneous Learning Models with Boosting. In: Leung C.S., Lee M., Chan J.H. (eds) Neural Information Processing. ICONIP 2009. LNCS, vol 5863. Springer, Berlin, Heidelberg. Google ScholarDigital Library
- Witten, I. H., Frank, E., et al. 2011. Data mining: Practical machine learning tools and techniques. New York: Elsevier. Google ScholarDigital Library
- Jose V R R, Winkler R L. 2008. Simple robust averages of fore-casts: some empirical results. Int J Forecast, 24(1), 163--169.Google ScholarCross Ref
- Armstrong J S. 2001. Principles of forecasting: a handbook for researchers and practitioners. Academic Publishers, Norwell, MA.Google Scholar
- Nascimento D.S.C., Coelho A.L.V. 2009. Ensembling Heterogeneous Learning Models with Boosting. In: Leung C.S., Lee M., Chan J.H. (eds) Neural Information Processing. ICONIP 2009. Lecture Notes in Computer Science, 5863. Springer, Berlin, Heidelberg. Google ScholarDigital Library
- Liu, X., Lu, F., Zhang, H. and Qiu P. 2013. Intersection delay estimation from floating car data via principal curves: a case study on Beijing's road network. Frontiers of Earth Science, 7(2), 206--216.Google ScholarCross Ref
Index Terms
- SHE: Stepwise Heterogeneous Ensemble Method for Citywide Traffic Analysis
Recommendations
Skype-Hunter: A real-time system for the detection and classification of Skype traffic
In the previous years, Skype has gained more and more popularity, since it is seen as the best VoIP software with good quality of sound, ease of use and one that works everywhere and with every OS. Because of its great diffusion, both the operators and ...
Sensitivity analysis for predictor variables in the MSAE regression
The minimum sum of absolute errors MSAE regression is more resistant to outliers, than the least squares regression, in the values of the response variable and long-tailed error distributions. Because all observations are used to compute the least ...
A Unifying Framework for Learning the Linear Combiners for Classifier Ensembles
ICPR '10: Proceedings of the 2010 20th International Conference on Pattern RecognitionFor classifier ensembles, an effective combination method is to combine the outputs of each classifier using a linearly weighted combination rule. There are multiple ways to linearly combine classifier outputs and it is beneficial to analyze them as a ...
Comments