Create a Forecasting Profile

This topic explains how you can create a forecasting profile for a demand, demand and supply, or replenishment plan.

A forecasting profile is a collection of features, forecasting methods, and forecasting parameters for use during the generation of demand forecasts.

Follow these steps to create a forecasting profile:

1. In the Supply Chain Planning work area, for a demand, demand and supply, or replenishment plan, in the More Actions menu, select View More.
   The Actions drawer will open.
2. Under Configuration, select Forecasting Profiles.
   The Forecasting Profiles page will open on a new tab of your web browser.
3. Select Create.
   The guided process for creating forecasting profiles will open on the same tab.
4. In the General step, do the following:
   1. Enter the forecasting profile name.
   2. In Machine Learning Type, select Bayesian if you want the forecasting profile to be based on Bayesian machine learning and use the algorithms in Oracle Demand Management.
   3. Enter a description for the forecasting profile.
   4. In Enable in Plan Type, select the plan types for which the forecasting profile can be selected in the guided process for creating or editing a plan.
   5. In Forecasting Table, select the forecasting table that will be used for building the forecast tree for the forecasting profile.
      The forecasting table defines the data-aggregation levels that are used in the forecasts. For information, see Overview of Forecast Trees.
   6. In Input Measure, select the input measure for the forecasting profile.
      This measure's data is used as the basis of historical demand in forecasts.

      The measures available for selection are in the selected forecasting table and dimensioned by time.

   7. In Output Measure, select the output measure for the forecasting profile.
      This measure stores the forecast after the plan is run.

      The measures available for selection are in the selected forecasting table and dimensioned by time. The output measure must be stored and can't be the same as the input measure.

   8. In Measure Catalogs, select all the user-defined measure catalogs that you expect are required for plans that use the forecasting profile.
      Measures that are automatically created by the forecasting profile will be added to the selected user-defined measure catalogs. Predefined measure catalogs aren't available for selection.
   9. Select Continue to move to the Features step of the guided process.
5. In the Features step, configure the feature groups and features for your forecasting profile.
   If your forecasting profile uses the Bayesian machine learning type, for information, see the topic on setting up features for Bayesian machine learning.

   Select Continue to move to the Methods step of the guided process.

6. In the Methods step, configure the forecasting methods for your forecasting profile.
   If your forecasting profile uses the Bayesian machine learning type, for information, see Forecasting Methods.

   Select Continue to move to the Parameters step of the guided process.

7. In the Parameters step, configure the forecasting parameters for your forecasting profile.
   If your forecasting profile uses the Bayesian machine learning type, for information, see Forecasting Parameters.
8. Select Submit to close the guided process and create the forecasting profile.
   A confirmation message will be displayed, and the newly created forecasting profile will be visible on the Forecasting Profiles page.