Forecast Confidence Index (FCI) is an indicator of statistical confidence in a particular unit forecast value based on the amount of historical information that is used to generate the forecast.
The index values for FCI range from 0 (zero) to 1 (one). An FCI of 1 indicates high confidence in the forecast value based on the historical information, while an FCI of 0 indicates high uncertainty.
Caution
FCI should not be interpreted as a measure of the absolute accuracy for the unit forecast value in comparison to the actual value. It should only be used as a measure of the amount as well as of the quality for the statistical information in order to generate the forecast value. Thus, when the FCI is 1 (or high), it should not be expected that the actual value would be identical to the forecast value, as there would be normal statistical fluctuations and other unpredictable factors. To determine the expected statistical confidence range for a forecast, other indicators such as expected mean absolute percent error (MAPE), low unit forecast, and high unit forecast should be used.
You can define the FCI values for low, medium and high ranges in Customizing for Cross-Application Components by choosing .
This categorization is then used for the analysis within promotion offer forecasting. The system displays the defined FCI values by default and allows you to override the forecast as required.
Available as of Release | BI Content 706 SP01 |
|---|---|
Aggregation | Summation |
Exception Aggregation | Average |
Calculation | No |
Restriction | No |
Object Type | Name | Technical Name |
|---|---|---|
InfoObject | Calendar Day | 0CALDAY |