Multi-instance Support Vector Machine Based on Convez Combination

摘要：

This paper presents a new formulation of multiinstance learning as maximum margin problem,which is an extension of the standard C-support vector classification.For linear classi fication,this extension leads to,instead of a mixed integer quadratic programming,a continuous optimization problem,where the objective function is convex quadratic and the constraints are ei ther linear or bilinear. This optimization problem is solved by an iterative strategy solving a convex quadratic programming and a linear programming alternatively.For non-linear classification,the corresponding iterative strategy is also established,where the kernel is introduced and the related dual problems are solved. The preliminary numerical experiments show that our approach is com petitive with the others.

关键词： Multi-instance classification Support vector machine Convez combination

作者: Zhi-Xia Yang Naiyang Deng

作者单位: College of Mathematics and System Science,Xinjiang University,Urumuchi 830046 Academy of Mathematics College of Science,China Agricultural University,100083,Beijing,China

会议类型: 国际会议

会议名称: The 8th International Symposium on Operations Research and Its Applications(第八届运筹及其应用国际专题讨论会 ISORA09)

会议地点: 张家界

会议语种:英文

页码: 481-487

在线出版日期: 2009-09-20（万方平台首次上网日期，不代表论文的发表时间）