Hi,
I am trying to do multi step ahead time series forecasting by using SVR. When I do forecast for training dataset, the results are pretty good. However, when I do iterative forecasting by adding last prediction values to the training dataset, the predictions get lower and lower values while there is no such pattern in the test set. How can I fix this problem and how should I do multi ahead forecasting with SVR in general? I am using (at least trying to use) Recursive Multi Step Ahead Forecast Strategy. Any help would be appreciated!
Dear Suzgun,
Kindly refer to the paper below.
Helmi, A., Fakhr, M. W., & Atiya, A. F. (2018). Multi-step ahead time series forecasting via sparse coding and dictionary based techniques. Applied Soft Computing, 69, 464-474.
From them, I am using recursive multi step ahead forecasting strategy. But my predictions have completely different pattern than the training set. I do not know how to fix it or if i am applying it in a wring way or not.
Dear Suzgun,
Kindly refer to the paper below.
Helmi, A., Fakhr, M. W., & Atiya, A. F. (2018). Multi-step ahead time series forecasting via sparse coding and dictionary based techniques. Applied Soft Computing, 69, 464-474.
- Badges
- Science topic