SAP IBP for Demand

Objective

After completing this lesson, you will be able to explain what's new for demand

SAP IBP Demand Forecasting 2408

The figure discusses the challenges and strategies for tuning parameters in intermittent forecast algorithms. It highlights the difficulty of parameter tuning due to missing values and the risk of overfitting in intermittent time series. It suggests optimizing Alpha and Beta coefficients in Croston TSB by considering characteristics like the Coefficient of Variation (CoV) and using a large pre-trained model for proposing coefficient values. Additionally, it notes that while optimized coefficients generally improve forecast accuracy, they may not always outperform default values, though significant improvements are seen with larger datasets.
The figure illustrates the configuration and execution of a forecast model in SAP, specifically using the Croston TSB method. It highlights that coefficient calculation is enabled by default for newly created forecast models. The interface shows settings for alpha and beta coefficients, with an example where the calculated alpha is 0.1921797985641155 and the calculated beta is 0.06387471918957909. The bottom table confirms the execution of the Croston TSB method for the product ID prod01 in the planning area 2408TSBopt, with the calculated coefficients displayed in the message column.

When coefficient calculation is disabled, values specified on the forecast model will be used for all planning objects.

By enabling the calculation of coefficients, IBP can fine-tune the parameters for each planning object separately.

Log in to track your progress & complete quizzes