R: tune.svm return same performance for every combination of parameter -


i have dataset 22 subjects , ~20000 features. want train supporting vector regression nu model linear kernel. in order chose best parameter model using tune function in package e1071 following call

out = tune(svm,dependentvariable~.,data=data,kernel="linear", ranges=list(cost=10^(-1:2), nu=seq(.1,1,.1))). 

the output of command is, expected, error , dispersion each combination of parameters, problem same error , dispersion each combination.

this strange, considering exploring entire possible parameter space nu.

what possibly going wrong ?


Comments

Popular posts from this blog

python - How to insert QWidgets in the middle of a Layout? -

python - serve multiple gunicorn django instances under nginx ubuntu -

module - Prestashop displayPaymentReturn hook url -