Releases
2.7.0
Highlights
Add EmbeddingSegmentTransform
Add EmbeddingWindowTransform
Add tutorial 210-embedding_models
about using embedding models
Add parameter drop_zero
into MRMRFeatureSelectionTransform
Allow RNNModel
, MLPModel
, DeepARNativeModel
, DeepStateModel
to work with categorical features
Allow encoders to return numeric features
Full changelog
Add TS2VecEmbeddingModel
model (#253 )
Add EmbeddingSegmentTransform
(#265 )
Add EmbeddingWindowTransform
(#265 )
Add TSTCCEmbeddingModel
(#294 )
Add 210-embedding_models
example notebook (#304 )
Add parameter drop_zero
into MRMRFeatureSelectionTransform
(#308 )
Changed
Allow RNNModel
to work with categorical features (#334 )
Allow DeepARNativeModel
and MLPModel
to work with categorical features (#336 )
Allow DeepState
to work with categorical features (#342 )
Allow encoders to return numeric features (#352 )
Enable cancelling old CI/CD runs after changes in a branch (#339 )
Fixed
Fix FordA download url in classification notebook (#309 )
Allow seaborn
dependency to have higher version (#319 )
Fix MRMRFeatureSelectionTransform
to correctly handle less-is-better relevance_table
(#308 )
Fix PatchTSModel
fails when using additional features (#376 )
Fix 101-get-started
notebook to be rendered correctly (#340 )
Fix DeepStateModel
forecasting problem with horizon=1 (#377 )
You can’t perform that action at this time.