AttributeError: 'SVC' object has no attribute 'best_estimator_'

Question

I'm trying to use GridSearchCV for SVM linear but I get this error:

AttributeError: 'SVC' object has no attribute 'best_estimator_'

the code of the SVM linear:

classifier = SVC()
classifier = GridSearchCV(classifier, {'C':[0.001, 0.01, 0.1, 1, 10,0.1, 100, 1000]}, cv=3, n_jobs=4)
classifier = SVC(kernel='linear')
classifier.fit(train_vectors, train_labels)
classifier = classifier.best_estimator_

can anyone help?

Answer 1

Do this:

classifier = SVC(kernel='linear')
gridsearch = GridSearchCV(classifier, {'C':[0.001, 0.01, 0.1, 1, 10,0.1, 100, 1000]}, cv=3, n_jobs=4)
gridsearch.fit(train_vectors, train_labels)

best_params = gridsearch.best_params_
classifier = gridsearch.best_estimator_