Orange Forum • View topic - Regression Tree

Regression Tree

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

Regression Tree

Postby collinan » Thu Sep 16, 2010 15:31

Hi Orangers!
I have a simple dataset (3 continous predictors + 1 discrete class response).
I did not manage to develop an individual Regression Tree classification (no-entry sign) while this same learner gave classification results when tested with other learners (Test Learners widget).
Could you please indicate me how to activate this learner, individually?
Many thanks,
Antoine.

Postby Ales » Fri Sep 17, 2010 10:15

The Regression tree classifier is used only for learning on data with continuous class. For data with discrete class use Tree learner in the Classify category.

Postby collinan » Wed Sep 22, 2010 16:05

Thanks Ales.
But could you enlighten me about the results stemmed from this regression tree while learning a discrete class?

Postby collinan » Thu Sep 23, 2010 19:30

Wether the Regression tree gave results with a discrete class response (within the Test Learner widget), where/how can I find the process (Tree) which provided these results?

Postby Ales » Fri Sep 24, 2010 9:43

But could you enlighten me about the results stemmed from this regression tree while learning a discrete class?


A mistake in Regression Tree learner widget enabled the resulting learner to default to a suitable attribute score measure for a discrete class.


Return to Questions & Support