Same cross-validation for multiple classifiers to have fair comparison?

[Am_HR7]

Hi guys,

I’m new to KNIME and I would like to use cross-validation for multiple classifiers on a dataset to have a fair comparison between them (give all classifiers the same training and test set). As the X-Aggregator node just has one input, I do not know how to connect multiple classifiers to it.

Thanks in advance!

[iperez]

Hi @Am_HR7. The X-Partitioner and X-Aggregator nodes should be paired, you can not connect multiple X-Aggregator nodes to a single X-partitioner one. In order to have the same sets for all models just use the same random seed across all the X-Paritioner nodes.

[Am_HR7]

Thanks!
So if I choose same Random seed for all the X-Partitioner nodes, I will get same training & test set?

[iperez]

Yes you get the same partition in all nodes.

[Am_HR7]

Thanks, Have a great day!