Time Series Forecasting On Etth2 96 2

MAE

MSE

評価結果

このベンチマークにおける各モデルのパフォーマンス結果

モデル名	MAE	MSE	Paper Title
FiLM	0.272	0.127	FiLM: Frequency improved Legendre Memory Model for Long-term Time Series Forecasting
PatchMixer	0.268	0.119	PatchMixer: A Patch-Mixing Architecture for Long-Term Time Series Forecasting
PatchTST/64	0.284	0.131	A Time Series is Worth 64 Words: Long-term Forecasting with Transformers
DLinear	0.279	0.131	Are Transformers Effective for Time Series Forecasting?
NLinear	0.278	0.129	Are Transformers Effective for Time Series Forecasting?
SegRNN	0.272	0.121	SegRNN: Segment Recurrent Neural Network for Long-Term Time Series Forecasting

0 of 6 row(s) selected.