I’d like to minimize the generalization error by pruning a decision tree using a validation set.
I know that Knime actually doesn’t support this feature. So I thought to configure a R Node like the R Learner, but with two input ports: one for the training set and one for the validation set. Is there any way to accomplish that?
Maybe the attached picture can explain better the question.
Somthing llike the picture in attachment should work.
It works, great!
Thank you very much, Fabienc!