Optimization of SVM Method with RBF Kernel

Abstract:

Usually there is no a uniform model to the choice of SVMs kernel function and its parameters for SVM. This paper presents a bilinear grid search method for the purpose of getting the parameter of SVM with RBF kernel, with the approach of combining grid search with bilinear search. Experiment results show that the proposed bilinear grid search has combined both the advantage of moderate training quantity by the bilinear search and of high predict accuracy by the grid search.

• Periodical: Applied Mechanics and Materials (Volumes 496-500)
• Pages: 2306-2310
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.496-500.2306
• Online since: January 2014
• Authors: Cheng Wang
• Keywords: Bilinear Grid Search, Model Selection, Parameter Optimization, RBF Kernel, Support Vector Machine (SVM)

References

[1] V N Vapnik: The Nature of Statical Learning Theory (Springer-Verlag, NY 1995).

[2] Zhang Xiaolong, Li Lin. Support Vector Machine and its Application Research. Computer Application Research. 2005. 12.

[3] Sarle W S. Periodic Posting to the Usenet Newsgroup Comp. Ai. Neural-nets. Neural Network FAQ. (1997).

[4] Keerthi S S, Lin C J. Asymptotic Behaviors of Support Vector Machines with Gaussian Kernel. Neural Computation. 2003. 15: 1667~1689.

DOI: 10.1162/089976603321891855

[5] Lin H T, Lin C J. A Study on Sigmoid Kernels for SVM and the Training of Non-PSD Kernels by SMO-type Methods. Technical report, (2003).

[6] O Chapelle, V Vapnik et al. Choosing multiple parameters for support vector machines. Machine Learning, 2002. 46: 131~159.

[7] Wang Peng, Zhu Xiaoyan. Model Selection of SVM with RBF Kernel and its Application. Computer Engineering and Applications. 2003. 24: 72~73.

[8] Information on http: /www. ics. uci. edu/~mlearn.