Options
Run forecast operator on STL decomposition
Learner I
Hello,
I'm a new user with Rapidminer and exploring some time series.
I have created a process with an STL decomposition and have 2 questions:
1) I see a clear seasonality pattern but the amplitude is growing (i.e. there is an increasing trend in my seasonality). Is this normal or should a look to set a parameter to try and keep the seasonality constant
2) Can I apply a forecast operator on the decomposition model to predict values in the future?
Thanks,
Bart
I'm a new user with Rapidminer and exploring some time series.
I have created a process with an STL decomposition and have 2 questions:
1) I see a clear seasonality pattern but the amplitude is growing (i.e. there is an increasing trend in my seasonality). Is this normal or should a look to set a parameter to try and keep the seasonality constant
2) Can I apply a forecast operator on the decomposition model to predict values in the future?
Thanks,
Bart
0
Answers
1) Yes, it's normal to have an increasing trend in the seasonal component of time series data.
2) Feel free to review this webinar briefing different time-series approaches to handle such variations.
https://rapidminer.com/resource/time-series-foundations/
Bonus: Here's an online book to understand Foresting Concepts!
https://otexts.com/fpp2/
Cheers,
Pavithra