I am a Knime noob and I might be mistaken but the K Nearest Neighbors nodes cannot be use to predict non classifiers such as a continuous ranking score between 1 and 5. Is that correct? I ask because rapidminer has an extension that does it. Is the only way around this is custom scripting?
Sorry for the delayed reply. I took a quick look and I believe you’re correct, the current implementation of KNN and K-means in KNIME only support classification and not regression. So you’d need to go the Python or R route in this particular case.