I've been trying to implement the classical Cat vs dogs classifier using sklearn.svm.SVC. However, I wasn't able to have more than 51% correct ratio (which is basically random, since there's only two classes). I tried varying gamma and C with no luck.

I started wondering if SVC is any good at all, for complex images.

I've used 400 x 300 grayscale images as input.

SVC is fine with hand-written digits but is it any good with more complex images ?

UPDATE : I just ran GridSearchCV to lookup good parameters for fit but it yielded Best parameters set found on development set:

{'C': 1, 'kernel': 'linear', 'gamma':0.001}

Which still gives me 0.51 ratio of good guess ...


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.