IBM Support

Definitions for MAPE and E2E Error

Question & Answer


Question

Definitions for MAPE and E2E Error

Answer

MAPE is the Median Absolute Percent Error of the model predictions compared to actuals during the modeling time period [(Actual - Forecast) / Actual].? MAPE is a measure of the forecast accuracy of the model itself, independent of optimization.

E2E Error is the percent error of the initial unit volume forecast in Price for the base E2E scenario compared to the actual historical volume for the same period as the scenario, only one year ago.? It takes into account not just how good the fit of the model was, but also how far off we are from reality once we've exported the coefficients and multipliers into DTPrice and run a scenario.? This is typically what most customers will view as forecast accuracy since it is the prediction they see in the application.

?(created by DT Support, last modified by DT Support on 2010-02-12T13:19:22.000Z)

50180000000TsITAA0
"

[{"Business Unit":{"code":"BU055","label":"Cognitive Applications"},"Product":{"code":"SSUP5C","label":"DemandTec"},"Component":"","Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"All Versions","Edition":"","Line of Business":{"code":"","label":""}}]

Document Information

Modified date:
08 December 2018

UID

ibm10770855