Information Technology Journal 1812-5638 1812-5646 Asian Network for Scientific Information 10.3923/itj.2013.2412.2418 Deng Weiping Zhou Jianzhong Zou Qiang Xiao Jian Zhang Yongchuan Hua Weihua 12 2013 12 12 Support vector machine is adopted in this paper to construct flood disaster evaluation model, which can be indicated as a comprehensively nonlinear classification issue. In this article, kernel function of SVM is optimized by kernel transformation and kernel parameters optimization. In order to discriminate the flood disaster evaluation indexes and really reflect their classification contributions, kernel function is weighted to promote classification performance and reduce the error influence by weak features. Further more, after analyzing the over learning issue of traditional grid search, an improved grid search is proposed to optimize the kernel parameters. The new search method illustrates it is more reasonable to make second search around some suboptimal solutions. By this search method, the final solution obtains high accuracy in testing samples and reduces the value of penalty factor. The experiment results show the promotion of classification precision by the dual optimization model and identify it could be a good choice for other classification issues.

]]> Adankon, M.M. and M. Cheriet, 2010 Neural Comput. Appl., 19 1197 1206 Chang, C.C. and C.J. Lin, 2007 LIBSVM-A library for support vector machines. 2007 Chen, T., 2011 Appl. Res. Comput., 28 139 141 Cristianini, N. and J. Shawe-Taylor, 2000 An Introduction to Support Vector Machines. Yuxia, H. and Z. Hongtao, 2012 Phys. Procedia, 25 588 594 Jiang, H.Y., S.L. Jing, T.Y. Chai and X.J. Zhou, 2010 J. Northeastern Univ., 31 494 497 Jiang, H.Y. and J.M. Liu, 2012 J. Inf. Comput. Sci., 9 5927 5934 Jiang, W., L. Deng, L. Chen, J. Wu and J. Li, 2009 Prog. Natural Sci., 19 1419 1425 Li, Q., J.Z. Zhou, D.H. Liu and X.W. Jiang, 2012 Safety Sci., 50 1275 1283 Lin, S.W., K.C. Ying, S.C. Chen and Z.J. Lee, 2008 Expert Syst. Appl., 35 1817 1824 Kapp, M.N., R. Sabourin and P. Maupin, 2012 Applied Soft Comput., 12 2550 2565 Ren, H. and M.D. Huo, 2009 Appl. Res. Comput., 3 867 869 Shi. M.D. and C.C. Yang, 2008 Expert Syst. Appl., 35 531 541 Vapnik, V.N., 2000 The Nature of Statistical Learning Theory. 2nd Edn., Wang, J., H.Y. Du, X.J. Yao and Z. Hu, 2007 Analytica Chimica Acta, 601 156 163 Wang, J.F., L. Zhang, G.X. Chen and X.W. He, 2012 Applied Sci. Technol., 6 28 31 Wang, T.H., S.F. Tian and H.K. Huang, 2009 J. Elect. Inf. Technol., 31 514 518 Wang, Y., S. Chen and H. Xue, 2011 Expert Syst. Appl., 38 12506 12513 Zhang, J.Y., S.L. Liu and Y. Wang, 2008 Prog. Natural Sci., 18 741 750