I have month-wise data that I'm using to forecast in Forecast Studio.
If I set up events (Pulse) for each month (of historical data) and apply the 'Use if Significant' filter, I'm getting substantially different results compared to not using events and just using the usual seasonality parameters.
Basically, then it looks like the events are picking up additional variation compared to what the seasonality alone picks up.
I'm wondering if anyone would have any insights to why this may be happening? Any thoughts greatly appreciated!
Thanks.
Seasonal components in time series models are not equivalent to dummy variables for the months, as they are often in linear regression models. Therefore, it is not suprising that you are getting different results.
For example, the following page details the equations for the additive Holt-Winters model, which is an exponential smoothing model with an additive seasonal component.
As you can see, the seasonal component S_t is dynamic, in that it depends on the past values S_(t-p), where p is the period.
In a regression model with seasonal dummies, the seasonal effect is a static additive component which does not vary with time.
Seasonal components in time series models are not equivalent to dummy variables for the months, as they are often in linear regression models. Therefore, it is not suprising that you are getting different results.
For example, the following page details the equations for the additive Holt-Winters model, which is an exponential smoothing model with an additive seasonal component.
As you can see, the seasonal component S_t is dynamic, in that it depends on the past values S_(t-p), where p is the period.
In a regression model with seasonal dummies, the seasonal effect is a static additive component which does not vary with time.
Registration is now open for SAS Innovate 2025 , our biggest and most exciting global event of the year! Join us in Orlando, FL, May 6-9.
Sign up by Dec. 31 to get the 2024 rate of just $495.
Register now!
Learn how to run multiple linear regression models with and without interactions, presented by SAS user Alex Chaplin.
Find more tutorials on the SAS Users YouTube channel.