What is the impurity criterion and minimum impurity required in Tree Ensemble (Regression) node in KNIME

[Anushka]

Hi All,

I am using Tree Ensemble (Regression) node for my modelling exercise. I am interested to know what’s the impurity criterion used to measure the quality of split and minimum value required for the criteria (Something similar is given by Python’s RandomForestRegressor function)
I was not able to find any reference for the criterion in the documentation too. It would be great if anyone could help me with that

Thanks,
Anushka

[marten_kose]

Hi @Anushka,

for the Tree Ensemble Regression Learner the “impurity criterion” is mean squared error. Definitely something to be improved in the documentation, thanks for pointing that out.

[marten_kose]

With version 4.3.4 the node description has been adjusted and now provides additional information on the impurity criterion.