How You Enable Nodal Tuning for Forecasting Nodes

This topic explains how you can configure the forecasting parameters in your user-defined forecasting profile of the Bayesian machine-learning type to enable nodal tuning for the forecasting nodes (combinations) in your demand or demand and supply plan.

When you enable nodal tuning for your plan, if hyperparameter tuning is successful for a forecasting node in a plan run, then the tuned settings are automatically applied to the forecasting node in later plan runs until the value of the HypertuneHalfLife forecasting parameter is reached.

These are the use cases for nodal tuning:

• You perform hyperparameter tuning periodically rather than on an ongoing basis.

  For example, you run hyperparameter tuning once every six months with the HypertuneSamplePercent forecasting parameter set to 100%, you run your demand plan once a week, and you want to use the tuned forecasting parameters in your demand plan.

• You want to perform incremental tuning.

  You've changed the values for hyperparameter tuning after it was last performed, and you want the previously tuned settings to be merged with the presently tuned settings. The values for previously tuned forecasting parameters that aren't in the current set of tuned forecasting parameters are retained in the tuned-settings measure and used. For tuned forecasting parameters that are changed, the old values are overwritten with new values in the tuned-settings measure.

Use these forecasting parameters in the Nodal Tuning category to configure nodal tuning:

• EnableNodalTuning: Use this forecasting parameter to enable nodal tuning for the forecasting nodes as follows:

  If EnableNodalTuning Is	And HypertuneSamplePercent Is	And HypertuneOutputMode Is	Then
  0 (zero)	Not applicable	Not applicable	Nodal tuning isn't performed.
  1	0 (zero)	Not applicable	The previously tuned settings in the tuned-settings measure are automatically applied to the forecasting nodes.
  1	greater than 0 (zero)	2	The previously tuned settings for forecasting nodes are used, the tuned forecast is written to the Bookings Forecast: Tuned, Shipments Forecast Tuned, or <user-defined output measure>: Tuned measure, and the forecast according to the forecasting profile's forecasting parameters is stored in the output measure. Incremental tuning is performed.
  2	greater than 0 (zero)	1	The previously tuned settings for forecasting nodes are used, and the tuned forecast is written to the output measure for the forecasting profile. Incremental tuning is performed.
  2	greater than 0 (zero)	2	Nodal tuning isn't performed.

  The default value for Enable Nodal Tuning is 2.

• SetNodalTuningSource: Use this forecasting parameter to specify the measure that has the settings for nodal tuning.

  When you enter 0 (zero), the Bookings Forecast: Tuned Settings, Shipments Forecast: Tuned Settings, or <user-defined output measure>: Tuned Settings measure is used.

  When you enter 1, a selected, user-defined measure is used.

  You can use this forecasting parameter only when the value of the EnableNodalTuning forecasting parameter is 1 or 2.

  The default value is 0.

  Note: This forecasting parameter isn't presently supported and shouldn't be used.