H2O RF Regression and Gradient Boosting Feature Selection Loop

I created a Feature Selection Loop and attached it to the Feature Selection Filter using the KNIME Random Forest Regression node (graphic below).

I want to recreate this process using the H2O Random Forest and separately the H2O Gradient Boosting Machine Learner but cannot figure out how to assemble the various H2O nodes and the KNIME nodes to create a functional Feature Selection Loop.

Please provide your suggestions, thanks

Maybe it helps if you share your workflow. I assume you can use the feature selection loop (standard) then use Table to H2O node and then do the partioning and GBM using the specific H2O nodes and then bring the H2O back to normal KNIME table before the loop ends

@smithcreed here is a collection of feature selection nodes where you could switch out the model to H2O - I have not checked if and how they use loops

And then there is this about feature selection heat I wanted to hint at

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.