Long-term forecasting on Weather (MSE, MAE, Information Gain Per Cost)
0.222MSETSCOMP
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| TSCOMPModel category=AutoML2026.05 | 0.222 | 0.256 | — | |
| MomentModel category=LTSM2026.05 | 0.228 | 0.268 | — | |
| TimeFuseModel category=AutoML, Evaluation protocol=Zero-shot2026.05 | 0.233 | 0.27 | — | |
| TimeFuseModel category=AutoML, Evaluation protocol=Few-shot2026.05 | 0.236 | 0.27 | — | |
| AutoGluonModel category=AutoML2026.05 | 0.236 | 0.27 | — | |
| GPT4TSModel category=LTSM2026.05 | 0.236 | 0.271 | — | |
| FeDaLZero-shot=true, Forecasting Horizon=Avg., Total Param.#=28.42 M, Training Data=~ 231B2025.08 | 0.255 | 0.284 | 4.545 | |
| Time-MoEultraZero-shot=true, Forecasting Horizon=Avg., Total Param.#=2.4 B, Training Data=300B2025.08 | 0.256 | 0.288 | 4.061 | |
| Time-MoEbaseZero-shot=true, Forecasting Horizon=Avg., Total Param.#=113 M, Training Data=300B2025.08 | 0.265 | 0.297 | 8.605 | |
| Time-MoElargeZero-shot=true, Forecasting Horizon=Avg., Total Param.#=453 M, Training Data=300B2025.08 | 0.27 | 0.3 | 2.073 | |
| TimerModel category=LTSM2026.05 | 0.335 | 0.365 | — | |
| AutoTSModel category=AutoML2026.05 | 0.519 | 0.372 | — |