Time Series Forecasting on ETTm1 (5% train)
0.314MSETime-VLM
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Time-VLMforecast_horizon=96, training_data_ratio=5%2026.05 | 0.314 | 0.357 | |
| Time-LLMforecast_horizon=96, training_data_ratio=5%2026.05 | 0.316 | 0.377 | |
| SSDAforecast_horizon=96, training_data_ratio=5%2026.05 | 0.33 | 0.342 | |
| DLinearforecast_horizon=96, training_data_ratio=5%2026.05 | 0.332 | 0.374 | |
| FiCoTSInput sequence length=512, Predictive lengths={96, 192, 336, 720}, Training data ratio=5%, Method category=LLM-based2025.11 | 0.361 | 0.382 | |
| SSDARatio=5%2026.05 | 0.363 | 0.369 | |
| TimeVLMInput sequence length=512, Predictive lengths={96, 192, 336, 720}, Training data ratio=5%, Method category=LLM-based2025.11 | 0.364 | 0.385 | |
| Time-VLMRatio=5%2026.05 | 0.364 | 0.385 | |
| SSDAforecast_horizon=336, training_data_ratio=5%2026.05 | 0.369 | 0.373 | |
| Time-VLMforecast_horizon=336, training_data_ratio=5%2026.05 | 0.373 | 0.391 | |
| GPT4TSforecast_horizon=96, training_data_ratio=5%2026.05 | 0.386 | 0.405 | |
| PatchTSTforecast_horizon=96, training_data_ratio=5%2026.05 | 0.399 | 0.414 | |
| DLinearInput sequence length=512, Predictive lengths={96, 192, 336, 720}, Training data ratio=5%, Method category=Non-Transformer-based2025.11 | 0.4 | 0.417 | |
| DLinearRatio=5%2026.05 | 0.4 | 0.417 | |
| DLinearforecast_horizon=336, training_data_ratio=5%2026.05 | 0.402 | 0.416 | |
| Time-LLMRatio=5%2026.05 | 0.425 | 0.434 | |
| TIME-LLMParameters (M)=6.46, Memory (MiB)=3907, Eff.*MSE=0.39, Eff.*MAE=0.38, Few-shot (5% data)=true2026.03 | 0.43 | 0.44 | |
| VisionTSforecast_horizon=96, training_data_ratio=5%2026.05 | 0.444 | 0.433 | |
| Time-LLMforecast_horizon=336, training_data_ratio=5%2026.05 | 0.45 | 0.424 | |
| S^2IP-LLMInput sequence length=512, Predictive lengths={96, 192, 336, 720}, Training data ratio=5%, Method category=LLM-based2025.11 | 0.455 | 0.446 | |
| GPT4TSParameters (M)=11.7, Memory (MiB)=371.0, Eff.*MSE=0.20, Eff.*MAE=0.20, Few-shot (5% data)=true2026.03 | 0.47 | 0.45 | |
| GPT4TSInput sequence length=512, Predictive lengths={96, 192, 336, 720}, Training data ratio=5%, Method category=LLM-based2025.11 | 0.472 | 0.45 | |
| GPT4TSRatio=5%2026.05 | 0.472 | 0.45 | |
| TimeLLMInput sequence length=512, Predictive lengths={96, 192, 336, 720}, Training data ratio=5%, Method category=LLM-based2025.11 | 0.477 | 0.451 | |
| GPT4TSforecast_horizon=336, training_data_ratio=5%2026.05 | 0.485 | 0.459 | |
| PatchTSTforecast_horizon=336, training_data_ratio=5%2026.05 | 0.499 | 0.467 | |
| One-for-AllParameters (M)=0.55, Memory (MiB)=2.2, Eff.*MSE=3.92, Eff.*MAE=4.17, Few-shot (5% data)=true2026.03 | 0.52 | 0.47 | |
| PatchTSTInput sequence length=512, Predictive lengths={96, 192, 336, 720}, Training data ratio=5%, Method category=Transformer-based2025.11 | 0.526 | 0.476 | |
| PatchTSTRatio=5%2026.05 | 0.526 | 0.476 | |
| VisionTSRatio=5%2026.05 | 0.574 | 0.498 | |
| TimesNetforecast_horizon=96, training_data_ratio=5%2026.05 | 0.606 | 0.518 | |
| FEDformerforecast_horizon=96, training_data_ratio=5%2026.05 | 0.628 | 0.544 | |
| VisionTSforecast_horizon=336, training_data_ratio=5%2026.05 | 0.653 | 0.53 | |
| TimesNetInput sequence length=512, Predictive lengths={96, 192, 336, 720}, Training data ratio=5%, Method category=Non-Transformer-based2025.11 | 0.717 | 0.561 | |
| TimesNetRatio=5%2026.05 | 0.717 | 0.561 | |
| TimesNetParameters (M)=0.63, Memory (MiB)=2.9, Eff.*MSE=2.89, Eff.*MAE=3.39, Few-shot (5% data)=true2026.03 | 0.72 | 0.56 | |
| FEDformerInput sequence length=512, Predictive lengths={96, 192, 336, 720}, Training data ratio=5%, Method category=Transformer-based2025.11 | 0.73 | 0.592 | |
| EFDformerParameters (M)=16.8, Memory (MiB)=87.8, Eff.*MSE=0.12, Eff.*MAE=0.13, Few-shot (5% data)=true2026.03 | 0.73 | 0.59 | |
| FEDformerRatio=5%2026.05 | 0.73 | 0.592 | |
| iTransformerInput sequence length=512, Predictive lengths={96, 192, 336, 720}, Training data ratio=5%, Method category=Transformer-based2025.11 | 0.784 | 0.596 | |
| TimesNetforecast_horizon=336, training_data_ratio=5%2026.05 | 0.786 | 0.597 | |
| AutoformerInput sequence length=512, Predictive lengths={96, 192, 336, 720}, Training data ratio=5%, Method category=Transformer-based2025.11 | 0.796 | 0.62 | |
| AutoformerParameters (M)=10.53, Memory (MiB)=62.6, Eff.*MSE=0.18, Eff.*MAE=0.20, Few-shot (5% data)=true2026.03 | 0.8 | 0.49 | |
| FEDformerforecast_horizon=336, training_data_ratio=5%2026.05 | 0.807 | 0.628 | |
| TStationaryParameters (M)=2.02, Memory (MiB)=13.2, Eff.*MSE=0.84, Eff.*MAE=1.03, Few-shot (5% data)=true2026.03 | 0.86 | 0.6 | |
| ETSformerInput sequence length=512, Predictive lengths={96, 192, 336, 720}, Training data ratio=5%, Method category=Transformer-based2025.11 | 1.125 | 0.782 | |
| ETSformerParameters (M)=5.28, Memory (MiB)=31.4, Eff.*MSE=0.24, Eff.*MAE=0.29, Few-shot (5% data)=true2026.03 | 1.13 | 0.79 | |
| Informerforecast_horizon=96, training_data_ratio=5%2026.05 | 1.13 | 0.775 | |
| InformerParameters (M)=11.33, Memory (MiB)=65.8, Eff.*MSE=0.04, Eff.*MAE=0.08, Few-shot (5% data)=true2026.03 | 1.16 | 0.79 | |
| InformerRatio=5%2026.05 | 1.163 | 0.791 | |
| Informerforecast_horizon=336, training_data_ratio=5%2026.05 | 1.198 | 0.809 | |
| ReformerParameters (M)=5.80, Memory (MiB)=33.4, Eff.*MSE=0.09, Eff.*MAE=0.17, Few-shot (5% data)=true2026.03 | 1.27 | 0.83 |