Forecast Methods

On this page you can review all the forecast methods with their parameters set for each record in a separate table for each level.

Set up forecast method based on records' segmentation


Using segmentation combinations the user can decide to use manual setting for the data in the specific segment. This can be set up in the first widget after entering the page Forecast Methods The user can set up the methods for each ABC and XYZ pairs, but they will be used only when learning algorithm is turned off. This can be handled in the first column of the widget. If the user turns on/off the checkboxes that will automatically update learning algorithm usage for elements with the segmentation data.

Note: At that point if the user recalculates the forecast the latest learning algorithm results will be used.

To inherit settings from the segment table, the user should use the Refresh forecast settings widgets action in the top right corner of the Set up forecast methods based on redords' Rate of Sale(ABC) and Variability(XYZ) information table. Using this the proper settings will be set for the records.

Set up forecast methods for each record

../../_images/forecastmethodtable_high.jpg ../../_images/forecastmethodtable_mdeium.jpg ../../_images/forecastmethodtable_low.jpg

Near the forecast settings there are the ABC and XYZ classifications, with their parameters.

The segmentation identifiers are:

  • Tracking signal

  • ABC classification

  • XYZ classification

  • Percentage value in the total sold amount

  • Added percentage

  • Mean variance


The forecast method data can be edited.

To select multiple records,

  1. Go to the action menu in the top right corner of the table and select Set up methods collective.

  2. Use the checkboxes to select records.

  3. Click Set up forecast methods.

  4. Choose a forecast method to apply to selected records.


Recalculate forecast

  • Calculate Forecast for changed records: Using this main page action, the forecast will recalculated for those elements where the setup differs from the previous run settings. In this case learning algorithm is not used.

  • Calculate Forecast: The forecast will be recalculated for all records. Learning algorithm is used where is turned on.