I can't find a way to parametrize WEKA Learner Nodes - i.e. I'd like to optimize numTrees parameter in Random Forest Learner node by mens of a Parameter Optimization Loop.
There isn't really a way to do this right now, except by doing it within Weka as far as possible (e.g. through Weka Boosting - I don't think they have ParmOpt). Doing this in a more KNIME-ish, cleverly integrated way is high up on my wishlist as well, let's hope it'll still happen this year.
Ok, finally we have parametrized WEKA nodes, but it looks somehow strange, as for IBk the parameters are I, F, K, E, W, X, A. Two of them are of String type, rest are Unknown type... how to use it?