Title :
A new multi-class support vector machines
Author :
Dong, Xin ; Zhaohui, Wu ; Yunhe, Pan
Author_Institution :
Dept. of Comput. Sci. & Eng., Zhejiang Univ., Hangzhou, China
Abstract :
A new classification using support vectors is presented. Support vector machines that learn classification problem are specific to use hyperplane. We propose a novel approach that contains support vectors describing the hypersphere to separate the samples. We also generalize it in multi-class classification phrase. The experiment results on UCI datasets are presented
Keywords :
classification; learning automata; UCI datasets; classification; hyperplane; hypersphere; multiclass support vector machines; Computer science; Constraint optimization; Kernel; Lagrangian functions; Pattern recognition; Shape; Support vector machine classification; Support vector machines;
Conference_Titel :
Systems, Man, and Cybernetics, 2001 IEEE International Conference on
Conference_Location :
Tucson, AZ
Print_ISBN :
0-7803-7087-2
DOI :
10.1109/ICSMC.2001.973525
