Macro for Monotonically Constrained ML Models
Description
A macro for training machine learning binary classification models with monotonic constraints, both increasing and decreasing, on selected variables as well as unconstrained variables.
As an example, we can see the partial dependence plots of two variables with decreasing and increasing monotonic constraints against the same variables without constraint in red.
 
It also allows the use of trained models in scoring workflows, as the model is persisted to a user-selected location.
