Page 1 of 1

Apply SVM in P300 Classification

Posted: 28 Jan 2015, 07:22
by sandra2014
Hi everyone,
I have a problem about SVM in P300 Classification. I want to know what kind of kernel function mostly used in this field ? I have used SWLDA to classify P300 data which can gain adoptive result . While with the same feature vector , SVM get very poor performance . I use libsvm ,parameters are those : -s 3 -t 0 , after cross-validation get -c and -g . So I want to know what -t I should set ?

Thanks ,sincerely
Sandra .