In our fraud example we will use two APL functions:
- CREATE_MODEL_AND_TRAIN to create a classification model and train it against the
past claims
- APPLY_MODEL to calculate a fraud prediction score for each individual recent
claim
These two functions are represented with their inputs and outputs in the following
diagram

The APL functions don’t come preconfigured with the APL installation. They must be
created using the AFL system procedure SYS.AFLLANG_WRAPPER_PROCEDURE_CREATE.
