The Demand Forecasting extension provides functionality for generating demand forecasts for items based on history. It is using the Forecasting Model for Dynamics 365 Business Central that Microsoft created, also known as Time Series API.
The Demand Forecast Worksheet allows you to analyze the results of the different models using different periods for history before creating the Demand Forecast.
Once the desired model and periods are selected they can be saved against the individual Items or Stockkeeping Units, and the updates of the Demand Forecasts can be automated through the Job Queue.
The time series modules supported are as follows:
- ARIMA (autoregressive integrated moving average)
- ETS (Error, Trend, Seasonal)
- STL (Seasonal-Trend decomposition using LOESS)
- ETS+ARIMA (returns average as result)
- ETS+STL (returns average as result)
- TBATS (Exponential smoothing state space model with Box-Cox transformation, ARMA errors, Trend and Seasonal components)
- ALL (If you choose ALL the model compares the results of all models listed above and returns the one that has the lowest mean absolute percentage error [MAPE]).