What is holdout forecast in SAP CA-DDF-RT - Demand Data Foundation for Retail?

SAP Term: holdout forecast

• Component: CA-DDF-RT

• Component Name: Demand Data Foundation for Retail

• Description: A forecast performed for periods in the past for which actual data is already available but was not used in model estimation. Holdout forecasts are used to validate the configuration of forecasting functions as well as forecast accuracy. Example With a holdout period of 90 days, modeling will exclude the most recent 90 days of demand data when evaluating the demand influencing factors DIFs. This allows determining how accurately the forecast predicted the actual unit sales for the holdout period.

Smart SAP Assistant

• Key Concepts: 
  Holdout forecast is a forecasting technique used in the Demand Data Foundation for Retail (CA-DDF-RT) component of SAP. It is used to create a forecast based on a subset of historical data, which is then used to predict future demand. The holdout forecast is created by taking a portion of the historical data and using it to create a model that can be used to predict future demand. 
  
  How to use it: 
  To use the holdout forecast, first select the subset of historical data that will be used to create the model. This subset should be representative of the overall data set, and should include enough data points to create an accurate model. Once the subset has been selected, the model can be created using statistical methods such as regression analysis or time series analysis. Once the model has been created, it can be used to predict future demand. 
  
  Tips & Tricks: 
  When creating a holdout forecast, it is important to select a representative subset of data that accurately reflects the overall data set. Additionally, it is important to ensure that enough data points are included in the subset so that an accurate model can be created. Finally, it is important to use appropriate statistical methods when creating the model so that it accurately predicts future demand. 
  
  Related Information: 
  The holdout forecast technique is similar to other forecasting techniques such as time series analysis and regression analysis. Additionally, other components of SAP such as SAP Analytics Cloud also offer forecasting capabilities.