Replies: 2 comments
-
The command to get these is: from sktime.registry import all_estimators
all_estimators("forecaster", filter_tags={"scitype:y": ["both", "multivariate"]}, as_dataframe=True) If you discount composites where the tag indicates potential capability depending on components, the current forecasters in
I'll also move this to GitHub discussions as it is a Q&A issue. (feel free to follow up there, and mark as resolved if this has answered your question!) |
Beta Was this translation helpful? Give feedback.
-
On another note, you can also create pipelines that have make use of multivariate correlation by virtue of transformers applied. For instance, you could pipeline |
Beta Was this translation helpful? Give feedback.
-
Are there any forecasters in this package like VAR, which support multi-dimensional time series with inter-dimensional correlation?
Beta Was this translation helpful? Give feedback.
All reactions