Study of Predictive Method Based on SVM Optimal Model Selection

Abstract:

The computation time consuming and poor efficiency of prediction exist in the model selection of traditional SVM. By studing on kernel matrix, a SVM-based prediction method for selecting the optimal model framework SVR-D1.2 was proposed with the help of the kernel matrix’s symmetry and positive definition and kernel alignment. The method was applied to the prediction of wheat scab, and comparison experiments were done with the main existing methods. The result shows the method has more efficiency and precision of prediction in the occurrence tendency of wheat scab. Meanwhile, it is simple, practicable.

• Periodical: Applied Mechanics and Materials (Volume 65)
• Pages: 443-446
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.65.443
• Online since: June 2011
• Authors: Lch Gu, Zhw Ni, Zhj Wu
• Keywords: Feature Selection, Kernel Alignment, Support Vector Regression (SVR), Wheat Scab

References

[1] Chih P W, Chiu I T. Turning telecommunications call details to churn prediction: A data mining approach[J ] . Expert Systems with Applications, 2002 , 23(2) : 103 - 112.

DOI: 10.1016/s0957-4174(02)00030-1

[2] Kim H S, Yoon C H. Determinants of subscriber churn and customer loyalty in the Korean mobile telephony market [ J ] . Telecommunications Policy , 2004 , 28(9): 751 - 765.

DOI: 10.1016/j.telpol.2004.05.013

[3] Rosset S, Neumann E. Integrating customer value considerations into predictive modeling[C]. Third IEEE International Conference on Data Mining, 2003: 1 - 8.

DOI: 10.1109/icdm.2003.1250931

[4] Nath S V. Data warehousing and mining: Customer churn analysis in the wireless industry [D] . A thesis submitted to the faculty of the college of business in partial fulfillment of the requirements for the degree of master of business administration, May (2003).

[5] K. -R. Muller, A. Smola, G. Ratsch, B. Scholkopf, J. Kohlmorgen, and V. Vapnik. Predicting time series with support vector machines. In B. Scholkopf, C. J. C. Burges, and A. J. Smola, editors, Advances in Kerne Methods - Support Vector Learning, pages 243–254, Cambridge, MA, 1999. MIT Press.

DOI: 10.7551/mitpress/1130.003.0019

[6] Ruiz A, Lopez-de-Teruel P E. Nonlinear Kernel-based Statistical Pattern Analysis,. IEEE Transactions on Neural Networks [J], 2001, 12(1): 16-31.

DOI: 10.1109/72.896793

[7] Bernhard Schölkopf, John C. Platt, etc. Estimating the Support of a High-Dimensional Distribution, Neural Computation [J]. 2001, 13(7), 1443-1471.

DOI: 10.1162/089976601750264965

[8] Manevitz L, Yousef M. One-class SVMs for document classification[J]. Journal of Machine Learning Research, 2001, 2: 139-154.

[9] Yi Ming, Hui Wan, Lei Li , etal. Multi dimensional model based clustering for user behavior mining in telecommunications industry [C]. Proceeding of the Third International Conference on Machine Learning and Cybernetics, Shanghai, 2004: 26 - 29.

DOI: 10.1109/icmlc.2004.1382040

[10] Li xiangdong, Luo bin, etc. Optimal Model Selection for Support Vector Machines. Journal of Computer Research and Development[J], 2005, 42(4): 576-58.