This workflow shows how to use cross-validation in H2O using the KNIME H2O Nodes. In the example we use the H2O Random Forest to predict the multiclass response of the IRIS data set using 5-folds and evaluate the cross-validated performance.
This is a companion discussion topic for the original entry at https://kni.me/w/mm5DxEbEyrgDG6ug
Dear Knime Team,
In this workflow ‘H2O CV loop start’ first port is attached to both RF learner as well as the predictor node. I thought the second port of the start loop is test set and should go to the predictor node. Is this an error or I am not getting it right?
Many thanks in advance for clearing the doubt !
Good catch! I believe this workflow is incorrect for the reason you mentioned. I’ll see about updating it.
Edit: The original workflow, located here, has now been corrected. Thanks for pointing this out.