Long-term Time Series Forecasting on ETTh1 (MSE, MAE, DTW, TDI)
0.388MSELightGTS-mini
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| LightGTS-mini2025.08 | 0.388 | 0.419 | — | — | |
| FlowStateModel parameters=10.6M, Context length=4k, Zero-shot=true2025.08 | 0.393 | 0.403 | — | — | |
| STaTAveraged forecasting horizons T={96, 192, 336, 720}2026.05 | 0.401 | 0.417 | 6.797 | 0.065 | |
| Time-VLMAveraged forecasting horizons T={96, 192, 336, 720}2026.05 | 0.405 | 0.42 | 7.416 | 0.07 | |
| Time-LLMAveraged forecasting horizons T={96, 192, 336, 720}2026.05 | 0.408 | 0.424 | 7.273 | 0.086 | |
| PatchTSTAveraged forecasting horizons T={96, 192, 336, 720}2026.05 | 0.413 | 0.431 | 7.369 | 0.087 | |
| DLinearAveraged forecasting horizons T={96, 192, 336, 720}2026.05 | 0.423 | 0.437 | 7.258 | 0.084 | |
| GPT4TSAveraged forecasting horizons T={96, 192, 336, 720}2026.05 | 0.428 | 0.426 | 6.975 | 0.072 | |
| FEDformerAveraged forecasting horizons T={96, 192, 336, 720}2026.05 | 0.44 | 0.46 | 7.551 | 0.081 | |
| iTransformer2025.08 | 0.454 | 0.448 | — | — | |
| TimesNetAveraged forecasting horizons T={96, 192, 336, 720}2026.05 | 0.458 | 0.45 | 8.088 | 0.103 | |
| PatchTST2025.08 | 0.469 | 0.455 | — | — | |
| MOIRAI-L2025.08 | 0.51 | 0.469 | — | — | |
| TimeCMAAveraged forecasting horizons T={96, 192, 336, 720}2026.05 | 0.527 | 0.478 | 7.507 | 0.072 | |
| StationaryAveraged forecasting horizons T={96, 192, 336, 720}2026.05 | 0.57 | 0.537 | 8.721 | 0.123 |