The Partition platform recursively partitions data, automatically splitting the data at optimum points. The result is a decision tree that classifies each observation into a group. The classic example is turning a table of symptoms and diagnoses of a certain illness into a hierarchy of assessments to be evaluated on new patients.

The Neural platform implements a standard type of neural network. Neural nets are used to predict one or more response variables from a flexible network of functions of input variables. They can be very good predictors, and are useful when the underlying functional form of the response surface is not important.