摘要
文章对半监督三分类光滑支持向量机进行研究。首先,针对二叉树结构的分类顺序对分类精度影响的问题,提出采用基于类间相似方向数生成偏二叉树支持向量机的方法判别分类顺序。其次,将此判别方法应用于半监督三分类中,提出了基于类间相似方向数的半监督三分类光滑支持向量机算法。最后,进行五组数值试验。结果表明,与全监督三分类算法相比,本文算法得到的准确率相对要高,且能在已标记样本比例较少的情况下得到较高的准确率,从而表明了本文的半监督二叉树三分类光滑支持向量机具有优越性。
This paper makes a study on the semi-supervised tri-classification smooth support vector machine. Firstly,aiming at the influence of classification order of binary tree structure on classification accuracy,this paper proposes a method of generating partial binary tree support vector machine based on the number of similar directions between classes. Secondly,the paper applies this method to semi-supervised tri-classification to present a semi-supervised tri-classification smooth support vector machine algorithm based on the number of similar directions between classes. Finally,five sets of numerical experiments are carried out. The results show that compared with the fully supervised tri-classification algorithm,the proposed algorithm in the paper has relatively higher accuracy,and can get high accuracy when the proportion of marked samples is small,thus indicating that the proposed semi-supervised binary tree tri-classification smooth support vector machine has advantages.
引文
[1]Yuan Y.Forecasting the Movement Direction of Exchange Rate With Polynomial Smooth Support Vector Machine[J].Mathematical and Computer Modelling,2013,(57).
[2]Sun L,Xu J,Zhang X,et al.An Image Watermarking Scheme Using Arnold Transform and Fuzzy Smooth Support Vector Machine[J].Mathematical Problems in Engineering,2015.
[3]Wu Q,Zhang L,Wang W.New Family of Piecewise Smooth Support Vector Machine[J].Journal of Systems Engineering and Electronics,2015,26(3).
[4]Reddy I S,Shevade S,Murty M N.A Fast Quasi-Newton Method for Semi-supervised SVM[J].Pattern Recognition,2011,(44).
[5]Zhang X,Ma J,Li A,et al.Quintic Spline Smooth Semi-supervised Support Vector Classification Machine[J].Journal of Systems Engineering and Electronics,2015,26(3).
[6]张晓丹,马菁改.一个广义三次样条光滑半监督支持向量机[J].工程科学学报,2015,(3).
[7]王建建,何枫,吴子轩,陈丽莉.Padé32光滑支持向量机模型的构造及其应用[J].控制理论与应用,2018,(9).
[8]王建建,张晓丹.一族三次样条光滑半监督支持向量机[J].计算机应用与软件,2015,(8).
[9]王建建,何枫.基于类间相似方向数的二叉树支持向量机[J].统计与决策,2018,(4).
[10]张健沛,赵莹,杨静.最小二乘支持向量机的半监督学习算法[J].哈尔滨工程大学学报,2008,29(10).
[11]王安娜,李云路,赵锋云等.一种新的半监督直推式支持向量机分类算法[J].仪器仪表学报,2011,32(7).
[12]Du P,Tan K,Xing X.A Novel Binary Tree Support Vector Machine for Hyperspectral Remote Sensing Image Classification[J].Optics Communications,2012,(285).
[13]吕佳,邓乃扬,田英杰等.局部学习半监督多类分类机[J].系统工程理论与实践,2013,33(3).
[14]陶新民,曹盼东,宋少宇等.基于两阶段学习的半监督支持向量机分类算法[J].信息与控制,2012,41(1).
[15]Yuan Y,Fan W,Pu D.Spline Function Smooth Support Vector Machine for Classification[J].Journal of Industrial&Management Optimization,2007,3(3)./