Adjusting Statistical Forecasts: Utilizing different Methods for varying time horizons

BettinaLang
BettinaLang
in Modeling

Hello Anaplan Community,

currently we have a very simple statistical forecast tool in Anaplan. We can select between 15 forecast methods for a forecast category. We generate the statistical forecast using a single method for the next 5 years.

We aim to introduce a logic that enables selecting two distinct forecasting methods for a single category. For instance, we would like to use Simple Exponential Smoothing for the first six months and then switch to a moving average method thereafter.

Is there a straightforward way to implement this logic?

Looking forward to your ideas 😊

Best regards, Bettina

Tagged:

Answers

  • Unknown
    Unknown
    This content has been removed.
  • Prajjwal88
    Prajjwal88

    Hi @BettinaLang,

    Adding to @Erika383Rivera suggestion, You may also consider calculating forecast errors like MSE and MAPE for each method. Based on these error values, you could create a module to identify the Best-fit method automatically. Additionally, having an override option would allow users to manually select a preferred method if needed.

    Hope this helps!

Categories