Orange Forum • View topic - ROC curves for the same classifier on different datasets

ROC curves for the same classifier on different datasets

A place to ask questions about methods in Orange and how they are used and other general support.

ROC curves for the same classifier on different datasets

Postby re832003 » Sun Jun 30, 2013 5:18

For a classification problem, I have 2 separate datasets each of which has a different set of features. I want to use SVM for both datasets and get the roc curves to compare the performance of the features.

I used the wedgit and it's doing an aswesome job but not for my problem. When I add another data file, it considers it a separate test set which is not correct.

So how I can get the curves I want?

Thanks.

Re: ROC curves for the same classifier on different datasets

Postby Ales » Mon Jul 01, 2013 9:48

I am not entirely certain I understand your problem, but if you just want to run the same learner on two data sets simply use two 'Test Learner' widgets and connect them to the same SVM widget. However you will need to make sure you will have the same parameters selected in the 'Test Learner' widgets.

Re: ROC curves for the same classifier on different datasets

Postby re832003 » Mon Jul 01, 2013 11:03

2 datasets
same classifier
one ROC curve graph for the 2 datasets.


That's my problem.


Return to Questions & Support