Closed xiaotaofeng closed 1 year ago
this is my condition:linear kernel;feature is six;svm_cross_validation ; sample size is 10000 my questions: (1) why my project train is very slow? (2) can i need do preprocess?
If you use linear, please use liblinear instead of libsvm. Liblinear is much faster for linear kernel.
this is my condition:linear kernel;feature is six;svm_cross_validation ; sample size is 10000 my questions: (1) why my project train is very slow? (2) can i need do preprocess?