variable importance for support vector machine and navies Bayes classifiers in R

Question

[kavya.krishnamurthy284]

Hi all,

I’m working on building predictive classifiers in R on cancer dataset. I’m using random forest, support vector machine and navies Bayes classifiers. I’m unable to calculate variable importance on SVM and NB models

I end up receiving the subsequent error.

Error in UseMethod("varImp") : 
no applicable method for 'varImp' applied to an object of class "c('svm.formula', 'svm')"

I would greatly appreciate it if anyone could help me.

Thanks in advance.

Answer 1

[Jean-Karim Heriche]

It looks like you're using the caret package which has no implementation of varImp for SVM. As far as I know, caret can give variable importance only for algorithms that can do feature selection and the standard 2-norm SVM is not one of them. The only package that I know does feature selection for SVM is the package penalizedSVM. With the 1-norm SVM, the weights reflect the importance of the features.