Please read the following page of the Sklearn documentation.

The figure shown there (see below) illustrates why C should be scaled when using a SVM with 'l1' penalty, whereas it shouldn't be scaled C when using one with 'l2' penalty.

enter image description here

The scaling however does not change the scores of the models examined within the GridSearch. So what exactly is this scaling-step good for?


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.