计算机科学 ›› 2013, Vol. 40 ›› Issue (12): 104-107.
覃希,苏一丹,张雯
QIN Xi,SU Yi-dan and ZHANG Wen
摘要： 借助商空间框架下的粒度分析理论及其计算方法,提出将“粒度”的概念用于大规模SVM数据集的约减来建立商空间框架下的约减模型。该约减模型的约减方向是由远及近地向分类超平面削减,其削减幅度也伴随集合的缩小而由粗到细逐渐变化。同时,给出该模型的一种实现。实验证明,商空间框架下的SVM约减模型比普通SVM约减模型的压缩效果更好。
| [1] Rychetsky M,Ortmann S,Ullmann M,et al.Accelerated Training of Support Vector Machines[C]∥International Joint Conference on Neural Networks,1999．IEEE,1999:998-1003 [2] Hsu C-W,Lin C-J．A Comparison of Methods for MulticlassSupportVector Machines[J].Neural Networks,2002,13(2):415-425 [3] Yu H,Yang Jiong,Han Jia-wei．Classifying Large Data Sets Using SVMs with Hierarchical Clusters[C]∥Ninth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining,2003.New York:ACM,2003:306-315 [4] 奉国和,朱思铭.基于聚类的大样本支持向量机研究[J].计算机科学,2006,33(4):145-147 [5] Xu Yan-zi,Qin Hua.A New Optimazation Method of Large-scale SVMs Based on Kernel Distance clustering[C]∥Computational Intelligence and Software Engineering,2009．Wuhan:IEEE,2009 [6] Cevikalp H.New clustering algorithms for the support vectormachine based hierarchical classification[J]．Pattern Recognition Letters,2010,1(11):1285-1291 [7] He Qiang,Xie Zong-xia,Hu Qing-hua．Neighborhood basedsample and feature selection for SVM classification learning[J].Neurocomputing,2011,74(10):1585-1594 [8] 胡正平,高文.基于改进加权压缩近邻与最近边界规则SVM训练样本约减选择算法[J].燕山大学学报,2010,4(5):421-425 [9] 张钹,张铃.粒计算未来发展方向探讨[J]．重庆邮电大学学报:自然科学版,2010,2(5):538-540 [10] 郭庆文,王国胤,张清华．多粒度的图像检索方法研究[J]．计算机科学,2013,40(1):298-301 [11] 程伟,石扬,张燕平.粒度计算的三种主要方法[J]．计算机技术与发展,2007,7(3):91-94 [12] 李道国,苗夺谦,张红云.粒度计算的理论、模型与方法[J]．复旦学报:自然科学版,2004,4(5):837-841 [13] 张玲,张钹.模糊商空间理论[J].软件学报,2003,4(4):770-776 [14] 覃希,苏一丹.用双层减样法优化大规模SVM垃圾标签检测模型[J].计算机应用研究,2011,8(6):2095-2098 [15] 覃希,夏宁霞,苏一丹.基于支持向量机的垃圾标签检测模型[J].计算机应用研究,2010,7(10):3893-3895 |
| No related articles found! |
|
||
首页