Multiclass Core Vector Machine with Smaller Core Sets

摘要：

Traditional methods for solving multi-class problems, well-known as multi-SVMs, always combine certain decomposed binary-SVMs results to formulate the final decision function. The prevalent methods are one vs. one and one vs. all which are based on a voting scheme among the binary classi.ers to derive the winning class. However, they do not scale well with the data size and class number. Core Vector Machine (CVM) is a promising technique for scaling up a binary-SVM to handle large data sets with the greedy-expansion strategy, where the kernels are required to be normalized to ensure the equivalence between the kernel-induced spaces of SVM and Minimum Enclosing Ball (MEB). The idea proposed by CVM can also be utilized to formulate multi-SVM to MEB, by which we propose an approximate MEB algorithm with smaller core sets to handle multi-SVM. The experimental results on synthetic and benchmark data sets demonstrate the competitive performances of the method we proposed both on training time and training accuracy.

关键词： Support vector machines Minimum enclosing ball Core Sets Kernel methods Approximate algorithm

作者: Yongqing Wang Xiaotai Niu Liang Chang

作者单位: Department of Computer Science and Applications, ZhengZhou Institute of Aeronautical Industry Manage Department of Computer Science and Applications, ZhengZhou Institute of Aeronautical Industry Manage College of Information Science and Technology, Beijing Normal University, Beijing 100875, China Virt

会议类型: 国际会议

会议名称: The 22nd China Control and Decision Conference(2010年中国控制与决策会议)

会议地点: 徐州

会议语种:英文

页码: 525-530

在线出版日期: 2010-05-26（万方平台首次上网日期，不代表论文的发表时间）