Solving large-scale multiclass learning problems via an efficient support vector classifier  

作　　者：Zheng Shuibo Tang Houjun Han Zhengzhi Zhang Haoran 

机构地区：[1]School of Electrical and Information Engineering, Shanghai Jiaotong Univ., Shanghai 200030, R R. China [2]Dept. of Electronic Engineering, Zhejiang Normal Univ., Jinhua 321004, P. R. China

出　　处：《Journal of Systems Engineering and Electronics》2006年第4期910-915,共6页系统工程与电子技术（英文版）

摘　　要：Support vector machines （SVMs） are initially designed for binary classification. How to effectively extend them for multiclass classification is still an ongoing research topic. A multiclass classifier is constructed by combining SVM^light algorithm with directed acyclic graph SVM （DAGSVM） method, named DAGSVM^light A new method is proposed to select the working set which is identical to the working set selected by SVM^light approach. Experimental results indicate DAGSVM^light is competitive with DAGSMO. It is more suitable for practice use. It may be an especially useful tool for large-scale multiclass classification problems and lead to more widespread use of SVMs in the engineering community due to its good performance.Support vector machines （SVMs） are initially designed for binary classification. How to effectively extend them for multiclass classification is still an ongoing research topic. A multiclass classifier is constructed by combining SVM^light algorithm with directed acyclic graph SVM （DAGSVM） method, named DAGSVM^light A new method is proposed to select the working set which is identical to the working set selected by SVM^light approach. Experimental results indicate DAGSVM^light is competitive with DAGSMO. It is more suitable for practice use. It may be an especially useful tool for large-scale multiclass classification problems and lead to more widespread use of SVMs in the engineering community due to its good performance.

关 键 词：support vector machines （SVMs） multiclass classification decomposition method SVM^light sequential minimal optimization （SMO）. 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]