This content has been marked as final. Show 1 reply
I think you meant to type Classification instead of Clustering. Clustering is non supervised mining and is not tested.
For supervised mining functions, such as Classification and Regression, Data Miner utilizes a Build and Train approach.
The Build data is used to create the model.
The Train data is used to test the model.
If you do not pass in separate data flows to the build nodes, the Classification and Regression Build nodes will split the data for you based on you test settings.
The Classification goes a bit further and performs a stratified random split based on the target.