计算机应用与软件Issue(2)：138-141,4.DOI:10.3969/j.issn.1000-386x.2016.02.033
基于可变隶属度的模糊双支持向量机研究
RESEARCH ON OPTIONAL MEMBERSHIP-BASED FUZZY TWIN SUPPORT VECTOR MACHINE
摘要
Abstract
Twin support vector machine is a novel nonparallel binary classification algorithm,and its processing speed is much faster than the traditional support vector machines,but the twin support vector machine does not consider that different input sample points will have different contribution on optimal classification hyperplanes.When the distances between test points and two kinds of hyperplanes are equal in test phase,the twin support vector machine does not explicitly give the treatment approach for these equidistant points.In view of this,this paper proposes an optional membership-based fuzzy twin support vector machine.The membership of sample point closer to the class centre is determined by the distance between the point and the class centre,while the membership of sample point farther to the class centre is jointly determined by the distance of the point to class centre and the affinity of point.During the testing phase,if the equidistant points appear,they can be classified according to the equivalence ratio of equidistant points with various test points.Experimental results show that compared with the support vector machine,standard twin support vector machine,twin-bounder support vector machine and fuzzy twin support vector machine,this optional membership-based fuzzy twin support vector machine has highest classification accuracy.关键词
双支持向量机/支持向量机/等距点/等价性比例/模糊隶属度/分类Key words
Twin support vector machine/Support vector machine/Equidistant points/Equivalence proportion/Fuzzy membership/Classification分类
信息技术与安全科学引用本文复制引用
任建华,刘晓帅,孟祥福,王伟..基于可变隶属度的模糊双支持向量机研究[J].计算机应用与软件,2016,(2):138-141,4.基金项目
国家自然科学基金青年科学基金项目（61003162）；辽宁省教育厅项目（L2013131）。 （）
