文献类型：期刊文献

英文题名：Self-Universum support vector machine

作者：Liu, Dalian[1,2];Tian, Yingjie[3];Bie, Rongfang[4];Shi, Yong[3]

第一作者：Liu, Dalian

通讯作者：Tian, YJ[1]

机构：[1]Beijing Jiaotong Univ, Sch Comp & Informat Technol, Beijing 100044, Peoples R China;[2]Beijing Union Univ, Dept Basic Course Teaching, Beijing 100101, Peoples R China;[3]Chinese Acad Sci, Res Ctr Fictitious Econ & Data Sci, Beijing 100190, Peoples R China;[4]Beijing Normal Univ, Coll Informat Sci & Technol, Beijing 100875, Peoples R China

第一机构：Beijing Jiaotong Univ, Sch Comp & Informat Technol, Beijing 100044, Peoples R China

通讯机构：[1]corresponding author), Chinese Acad Sci, Res Ctr Fictitious Econ & Data Sci, Beijing 100190, Peoples R China.

年份：2014

卷号：18

期号：8

起止页码：1813-1819

外文期刊名：PERSONAL AND UBIQUITOUS COMPUTING

收录：;EI(收录号：20154801626982);Scopus(收录号：2-s2.0-85027940896);WOS:【SCI-EXPANDED(收录号:WOS:000344812200003)】；

基金：This work has been partially supported by grants from National Natural Science Foundation of China (Nos. 11271361, 71331005), Major International (Ragional) Joint Research Project (No. 71110107026), the Ministry of water resources' special funds for scientific research on public causes (No. 201301094).

语种：英文

外文关键词：Support vector machines; Twin support vector machines; Universum; Nonparallel; Classification

摘要：In this paper, for an improved twin support vector machine (TWSVM), we give it a theoretical explanation based on the concept of Universum and then name it Self-Universum support vector machine (SUSVM). For the binary classification problem, SUSVM takes the positive class and negative class as Universum separately to construct two classification problems with Universum; therefore, two nonparallel hyperplanes are derived. SUSVM has several improved advantages compared with TWSVMs. Furthermore, we improve SUSVM by formulating it as a pair of linear programming problems instead of quadratic programming problems (QPPs), which leads to the better generalization performance and less computational time. The effectiveness of the enhanced method is demonstrated by experimental results on several benchmark datasets.