I want to classify the significant features obtained from nonlinear methods using SVM-RBF kernel function. I have used 10-fold cross validation that divides 180 samples as training set and remaining 20 as testing set.

Can you please guide me through the steps on how to implement SVM with RBF kernel to classify the data into two groups?

  • $\begingroup$ Adding some details about the data and focusing on a particular step would help narrow this question, making it more accessible and more easily answered. $\endgroup$
    – whuber
    May 26, 2014 at 15:05
  • $\begingroup$ Did you consider feature extraction first? EEG data can't be loaded directly as features into SVM or any classifier. You need the signal processing part first. $\endgroup$
    – A.Rashad
    Oct 30, 2014 at 5:42

1 Answer 1


I'm not sure if my answer will be appropriate as description of your study above is very limited. I would recommend you using WEKA software that is free and LibSVM package. You can download the extension that will draw decision boundaries and I'd consider it as a best way to present results of your research and dependencies between features.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.