Competitive Forecast Properties
You can use the Competitive Method Properties to configure a competitive method and decide which forecast methods compete. Select the method with the lowest error that Error Analysis defines. This method is the winning method.
Note: This feature can calculate many forecast methods but can
require more time to perform the calculation. The competitive method is not a substitution for
informed decision-making about how to forecast. You can use the competitive method feature to
assist and guide the selection of a forecast method. You can also use the feature to
continually perform automatic selection between forecast methods. The winner changes based on
historical data when Error Analysis finds a forecast method with lower error.
You can access the dialog box from . Select Competitive Forecast Methods.
This table shows the property settings you can use:
| Setting | Description |
|---|---|
| General | This property includes Name, Category filter, Description, and Scope
(availability) information. The Category
field serves as a filter option in the Competitive Methods list. You can select an
existing category or specify a new one. Note: Scope
(availability) information is only available for administrators. For administrators,
select Public scope.
|
| Statistical Forecast Methods | This property provides a list of defined forecast methods so that users can select the method for the Competitive Forecast calculation. |
| Error Analysis | This property provides common and specialized forecast error methods. See Error Analysis. |
| Execution | This property defines the Forecast Horizon for a competitive forecast group. It
overrides the extrapolation settings of individual methods so that all methods in the
group forecast an equal number of periods. See Setting the Forecast Horizon for Competitive Forecast Execution. |
Note: This functionality is only enabled for administrators.