ECFP with Random Forest
Move onto doing random forest because SVM wasn't so good (at this case! not always!!). And I won't bother tuning parameters for SVM too.
File Reader: Read csv
Column Filter: Delete unnecessary columns
RDKit Fingerprint: Morgan FP as always
Random Forest Learner: default
Random Forest Predictor: default
Scorer: default
One thing I really like about RF learner
I can use fingerprint as-is!!!! This is so nice, should make this available for all learners.
Then what do I get...
Not bad at all!! RF is very easy to use. I probably can get around 0.8 by really tuning this? But wait, how about XGBoost, that was integrated into KNIME recently...