论文部分内容阅读
SVM handles classification problem only considering samples themselves and the classification effect depends on the characteristics of the training samples but not the current information of classified problem.From the phenomena of data crossing in system