Use Subsampling to Solve Imbalanced Dataset Problem for Automatic Incident Detection Algorithm

Abstract:

Considering the fact that the amount of traffic incident data is rare compared to the large amount of normal traffic state data in the real word, we proposed an Automatic Incident Detection (AID) algorithm based on subsampling method. First, an improved subsampling method based on Edited Nearest Neighbor Rule (ENN) algorithm was used to reconstruct the training set to get a balanced dataset. Then, the Support Vector Machine (SVM) was adopted as a classifier to detect traffic incidents. The real traffic data collected from the I-880 freeway in American was used to build the model and test the performance of the proposed AID algorithm. In addition, we made a comparison of the detection performances between the AID algorithm obtained by the original training set and the one by the relatively balanced training set. The experimental results show that the proposed AID algorithm based on subsampling is suitable for imbalanced dataset and can obtain a better detection performance.

• Periodical: Applied Mechanics and Materials (Volumes 587-589)
• Pages: 2114-2119
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.587-589.2114
• Online since: July 2014
• Authors: Miao Hua Li*, Shu Yan Chen
• Keywords: Automatic Traffic Incident Detection, ENN, Imbalanced Dataset, Subsampling, Support Vector Machine (SVM)

References

[1] Yuan, F. and R.L. Cheu. Transportation Research Part C: Emerging Technologies, 2003. 11(3-4): pp.309-328.

[2] Wang, F., Parallel Control and Management for Intelligent Transportation Systems: Concepts, Architectures, and Applications. IEEE Transactions on Intelligent Transportation Systems, 2010. 11(3): pp.630-638.

DOI: 10.1109/tits.2010.2060218

[3] Chen, S. and W. Wang. Expert Systems with Applications, 2009. 36(2): pp.4101-4105.

[4] R.L. Cheu, S.G. Transportation Research Part C – Emerging Technologies 3 (1995) 371–388.

[5] Y. Hawas. Transportation Research Part C 15 (2007) 69–95.

[6] Anguita, D., A. Boni and S. Ridella. IEEE Transactions on Neural Networks, 2003. 14(5): pp.993-1009.

[7] Chawla, N., N. Japkowicz and A. Kotcz, Editorial. ACM SIGKDD Explorations Newsletter, 2004. 6(1): pp.1-6.

DOI: 10.1145/1007730.1007733

[8] Wenchang Zheng , Shuyan Chen and Xuanqiang Wang. 2012 (11): page 58-62 +123 (In Chinese ).

[9] Laurikkala, J., Improving identification of difficult small classes by balancing class distribution. 2001: Springer.

[10] Batista, G., R. Prati and M. Monard. ACM SIGKDD Explorations Newsletter, 2004. 6(1): pp.20-29.

[11] Wilson, D.R., Martinez, T.R. Mach. Learn. 38 (2000) 257-286.

[12] CHEN, W., P. GUAN and Y. ZOU: School Of Electrical Southwest Jiaotong University, 2011. 46(1): page 63-67.

[13] Jian XIAO , Long Yu, Yifeng Bai. Learned journal of Southwest Jiaotong University, 2008, 43 (3): pages 297-303 (In Chinese ).

[14] Cortes, C. and V. Vapnik. 1995. 20(3): pp.273-297.

[15] MATLAB Chinese forum on SVM parameter c & g selected summary posts [matlab-libsvm]. http: /www. ilovematlab. cn/forum. php?mod=viewthread&tid=47819&fromuid=600178.

[16] Petty, K., Noeimi, H., Sanwal, K., Rydzewski, D., Skabardonis, A., Varaiya, P., et al. (1996). Transportation Research, 4C(3), 71–86.

DOI: 10.1016/0968-090x(96)00001-0

[17] Yuan, F. and R.L. Cheu. 2003. 11(3-4): pp.309-328.

[18] Gong Jiong. 2010, Southwest Jiaotong University, page 67.

[19] Turner, S. M., Albert, L., Gajewski, B., & Eisele, W. (2000). TRB (1719, p.77–84). Washington, DC: National Research Council.

[20] Wilson, D.L. IEEE Transactions on Systems, Man, and Cybernetics, 1972. 2(3): pp.408-421.