Metric: MAE (lower is better)
| # | Model↕ | MAE▲ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | NLinear | 0.177 | No | Are Transformers Effective for Time Series Forec... | 2022-05-26 | Code |
| 2 | FiLM | 0.178 | No | FiLM: Frequency improved Legendre Memory Model f... | 2022-05-18 | Code |
| 3 | PatchMixer | 0.179 | No | PatchMixer: A Patch-Mixing Architecture for Long... | 2023-10-01 | Code |
| 4 | SegRNN | 0.18 | No | SegRNN: Segment Recurrent Neural Network for Lon... | 2023-08-22 | Code |
| 5 | DLinear | 0.18 | No | Are Transformers Effective for Time Series Forec... | 2022-05-26 | Code |
| 6 | PatchTST/64 | 0.189 | No | A Time Series is Worth 64 Words: Long-term Forec... | 2022-11-27 | Code |