Title :
A Weighted Hyper-Sphere SVM
Author :
Zhang, Xinfeng ; Xu, Xiaozhao ; Cai, Yiheng ; Liu, Yaowei
Author_Institution :
Signal & Inf. Process. Lab., Beijing Univ. of Technol., Beijing, China
Abstract :
Generalized hyper-sphere SVM is a promising method for the pattern classification. The ratio of the support vectors from two classes of samples can not be adjusted conveniently by setting the parameters n and b in the generalized hyper-sphere SVM (GHSVM), which affects the generalization performance to some extent. A weighted hyper-sphere SVM is studied in this paper. The results shows that the margin may be obtained much more easily by weighted method rather than by adjusting the parameters n and b, which makes the classifier´s generalization performance much better than the original GHSVM.
Keywords :
generalisation (artificial intelligence); pattern classification; support vector machines; generalization performance; generalized hyper-sphere SVM; pattern classification; support vectors; weighted hyper-sphere SVM; Face detection; Image analysis; Image classification; Information processing; Pattern classification; Performance analysis; Signal processing; Support vector machine classification; Support vector machines; Tongue; Generalized hyper-sphere SVM; Weigheted; generalization performance;
Conference_Titel :
Natural Computation, 2009. ICNC '09. Fifth International Conference on
Conference_Location :
Tianjin
Print_ISBN :
978-0-7695-3736-8
DOI :
10.1109/ICNC.2009.398
