Question: Feature Selection for Regression Problems

[SymbolicSquared]

The examples provided apply Boruta for feature selection in classification problems. Can Boruta be accurately applied for feature selection in regression problems? If so, what regression estimator would be most appropriate? (i.e. RandomForestRegressor, GradientBoostingRegressor, etc.)

[Sandy4321]

yes, it would be great to answer on this question, since your wrote : Compatible with any ensemble method from scikit-learn ?

[danielhomola]

RandomForrestRegressor should work fine with Boruta.

[Sandy4321]

it is exactly the question: can it be used for Logistic regression from scikit learn or not?
from your answer fallows that not: it works only for RandomForrestRegressor from scikit learn
may you write clear way ; can it be used for Logistic regression from scikit learn???

[danielhomola]

It seems like you haven't read the paper or now how boruta works. It might be useful.

[Sandy4321]

Boruta conceptually should work with Logistic Regression , only the question is : did you coded this option?

[danielhomola]

No.

[Sandy4321]

interesting why?

[danielhomola]

because it's not what's in Boruta (you might have missed which package you're using) and besides that why would anyone use a linear classifier instead of RF or other ensemble?