Dimensionality Reduction Method of Training Sample Set for SVDD Based on Statistical Information

Abstract:

In order to solve problem of high time complexity of support vector data description (SVDD) in training process, a low complexity SVDD algorithm is proposed by introducing statistical information grid (STING).The algorithm applies STING division to sample set space using support vector distribution characteristics and kernel distances between samples and sphere’s centre. Based on position information of sample points, it rejects non-support vectors and obtained simplified sample set. The results show that proposed method can reduce training scale and time without decreasing classification accuracy.

• Periodical: Applied Mechanics and Materials (Volumes 220-223)
• Pages: 2097-2101
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.220-223.2097
• Online since: November 2012
• Authors: Mian Wu, Bin Chen, Bao Cheng Gao, Xiao Bin Cheng, Zhao Li Yan
• Keywords: Statistical Information Grid, Support Vector Data Description, Time Complexity

References

[1] D.M.J. Tax, R.P.W. Duin. Support vector domain decription. Pattern Recognition Letters, 1999, 20 (11-13):1191-1199.

DOI: 10.1016/s0167-8655(99)00087-2

[2] Bin Chen, Zhaoli Yan, Xiaobin Cheng. Machinery fault trend prediction based on SVDD and relative distance. Chinese Journal of Scientific Instrument, 2011, 32(7): 1558-1562.

[3] Changjun Chen, Yongzhao Zan, Chuanjun Wen. Support vector description discriminant analysis and its application to face recognition. Application Resrarch of Computers, 2010, 27 (2):488-490.

[4] Jing Xu, Ruiyin Shi, Yajiang Zhang. Model of IDS based on SVDD and cluster algorithm. Control and Decision, 2010, 25(3): 441-443.

[5] Chunhong Zheng, Licheng Jiao. Fuzzy pre-extracting method for support vector machine. Proceedings of the First International Conference on Machine Learning and Cybernetics, Beijing, China, 2002, 4:2026-2030.

DOI: 10.1109/icmlc.2002.1175393

[6] G.Fung, O.L. Mangasarian. Proximal suppoft vector machine classifiers. Proceedings KDD-2001: Knowledge Discovery and Data Mining, SanFrancisco,California, 2001: 77-86.

DOI: 10.1145/502512.502527

[7] Jigang Wang, Predrag Neskovic, Leon N Cooper. Selecting data for fast support vector machines training. Chapter 3 in Studies in Computation Intelligence, Springer. Verlag, 2007, 35: 61-84.

DOI: 10.1007/978-3-540-36122-0_3

[8] Sanyan Liu, Jinjin Liang, De Wu. Confidence support vector domain description. Journal of Systems Engineering and Electronics, 2009, 29 (4):852-857.

[9] Wei Wang, Jiong Yang, Richard Muntz. STING: A statistical information grid approach to data mining. Proceedings of the 23rd VLDB Conference Athens, Greece, 1997:186-195.