南京大学学报(自然科学版) ›› 2013, Vol. 49 ›› Issue (5): 644649.
张宇1,王文剑1,2,郭虎升1
Zhang Yu1, Wang Wenian1,2, Guo Hu-Sheng1
摘要: 粒度支持向量机(Granular support vector machine, GSVM)通过选取粒的代表点构成精简训练集以提高支持向量机(Support vector machine, SVM)的学习效率,然而选取个别代表点有可能丢失部分重要分类信息,导致模型泛化能力不高。针对这一问题,提出基于粒分布的GSVM(Distribution based GSVM, DGSVM)加速训练方法,该方法依照粒内正负样本分布选取粒代表点,根据粒的混合度将这些代表点分为精简训练样本集和修正集,使用精简训练样本集训练得到分类器,用修正集迭代优化分类器。本质上DGSVM是用少量难分的样本训练快速得到初始分类器,然后再进行进一步调整。在标准数据集上的实验结果表明DGSVM方法可以在保证算法学习效率的同时提高分类器的泛化能力。
[1] Vapnik V. Statistical learning theory. New York: Wiley, 1998, 493~520. [2] Platt J C. Sequential minimal optimization: A fast algorithm for training support vector machines. Technical Report, MSR-TR-98-14, 1999. [3] Keerthi S S, Shevade S K. SMO algorithm for least-squares SVM formulations. Neural Computation, 2003, 15(2): 487~507. [4] Zhang H Y,?Yi Y, Wu?J S. Network intrusion detection system based on incremental support vector machine. Contemporary Challenges and Solutions in Applied Artificial Intelligence Studies in Computational Intelligence, 2013, 489: 91~96. [5] Horng S J, Su M Y, Chen Y H, et al. A novel instrusion detection system based on hierarchical clustering and support vector machines. Expert Systems with Applications, 2011, 38(1): 306~313. [6] Wang W J, Guo H S, Jia Y F, et al. Granular support vector machine based on mixed measure. Neurocomputing, 2013, 101: 116~128. [7] Bai L F, Wang W J, Guo H S. A novel support vector machine active learning strategy. Journal of Nanjing University (Natural Sciences), 2012, 48(2): 172~181. (白龙飞,王文剑,郭虎升. 一种新的支持向量机主动学习策略. 南京大学学报(自然科学), 2012, 48(2): 172~181). [8] Ding S F, Qi B J. Research of granular support vector machine. Artificial Intelligence Review, 2012, 38: 1~7. [9] Tang Y C, Jin B, Zhang Y Q. Granular support vector machines with association rules mining for protein homology prediction. Artificial Intelligence in Medicion, 2005, 35: 121~134. [10] Tang Y C. Granular support vector machines based on granular computing, soft computing and statistical learning. Georgia Stage University: College of Arts and Sciences, 2006. [11] Liu Y, Song H Z. Study on constructing support vector machine with granular computing. Procedia Engineering, 2011, 15:3098~3102. [12] Liu X D, Luo B, Chen Z Q. Optimal model selection for support vector machines. Journal of Computer Research and Development, 2005, 42(2): 567~581.(刘向东骆斌陈兆乾. 支持向量机最优模型选择的研究. 计算机研究与发展2005, 42(4): 576~581) [13] Wang W J, Xu Z B, Lu V Z, et al. Determination of the spread parameter in the Gaussian kernel for classification and regression. Neurocomputing, 2003, 55(3-4): 643~663 [14] Liao S Z, Jia L. Constructing a new spherical kernel function. Journal of Computer Research and Development, 2007, 44(z2): 398~402.(廖士中贾磊一类新的球面核函数的构造. 计算机研究与发展2007, 44(z2): 398~402). [15] Wu T, He H G, He M K. Interpolation based Kernel function’s construction. Chinese Journal of Computers, 2003 ,26(8): 990~996. (吴涛贺汉根贺明科基于插值的核函数构造. 计算机学报2003, 26(8): 990~996). [16] UCI Machine Learning Repository. http://archive.ics.uci.edu/ml, 2009-10-16. |
No related articles found! |
|