Decision Tree node - weird outcome

Hello everybody,

Recently I executed a decision tree node twice, with minimum leaf size set to 30 and 60 respectively.
In other words, I executed the same workflow with 2 different minimum leaf sizes, one after the other.
And other settings remained the same.

The weird things was that with minimum leaf size= 30, there was no child nodes created in the result decision tree model.  As for minimum leaf size=60, there was 2 child nodes in the result. 30 is smaller than 60, and thus less restrictive, therefore under minimum leaf size=30, there should be equal or more child nodes created than under minimum leaf size=60.  But my result was the opposite.

Can anybody explain ?  Is there a bug in KNIME ? Is KNIME a flawed software ?
Thank you in advance. 

Marty

Hi Marty,

could you tell us more about the data you are using?

Best, Iris

PS: Could you please don't make double posts? Our users are very active however it might take more than the weekend before christmas to have an question answered