Splitting and Partitioning

Splitting has some features in common with partitioning, but the two are used in very different ways.

Partitioning divides the dataset randomly into either two or three parts: training, testing and (optionally) validation, and is used to test the performance of a single model.

Splitting divides the dataset into as many parts as there are possible values for a split field, and is used to build multiple models.

Partitioning and splitting operate completely independently of each other. You can choose either, both or neither in a modeling node.