ML task in multidimensional process (beta version)

You can add ML tasks to a multidimensional business workflow.

Tasks can be defined with advanced analytics linked to models either with parameters required at runtime or without parameters.

When launching a data processing, two check types are run:

• The input/output dataset cannot be partitioned by Business Cycle & Entities
• The type of the input/output dataset must be Reporting Layer.

If either of these is the case, the data processing fails and a message is shown:

By clicking on the “Details” button, you can display the details of the data processing explaining why the data processing failed.