关键词: 极端支持向量机, 支持向量机, 权

Abstract: Extreme Support Vector Machine（ESVM） is susceptibly interfered by noise and outliers, because it does not take into account the distribution of the sample points and gives the same punishment factor to all of the error terms of sample points. For this problem, Weight Extreme Support Vector Machine（WESVM） is proposed on the basis of ESVM. Because the distances from different types of samples to the center of the sample are different, the different weights are assigned to the different training samples. The experiment results show that WESVM has a better classified accuracy compared with ESVM and Extreme Learning Machine.

Key words: Extreme Support Vector Machine（ESVM）, Support Vector Machine（SVM）, weight