Weather Forecasting On Sd
Métriques
MSE (t+1)
MSE (t+6)
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | MSE (t+1) | MSE (t+6) |
|---|---|---|
| neural-dynamics-on-complex-networks | 0.5705 ± 0.0057 | 0.7414 ± 0.0294 |
| climate-modeling-with-neural-diffusion | 0.3561 ± 0.0055 | 0.7301 ± 0.0048 |
| climate-modeling-with-neural-advection | 0.1430 ± 0.0280 | 0.6516 ± 0.0657 |
| graph-networks-as-learnable-physics-engines | 0.6543 ± 0.1195 | 0.9872 ± 0.2425 |
| neural-dynamics-on-complex-networks | 0.5296 ± 0.0274 | 0.7542 ± 0.0730 |
| neural-dynamics-on-complex-networks | 0.5291 ± 0.0578 | 0.7862 ± 0.0475 |
| graph-networks-as-learnable-physics-engines | 0.7007 ± 0.0848 | 1.0422 ± 0.0673 |
| adaptive-graph-convolutional-recurrent | 0.2010 ± 0.0188 | 1.0181 ± 0.1275 |
| neural-dynamics-on-complex-networks | 0.5754 ± 0.0180 | 0.7954 ± 0.0110 |
| differentiable-physics-informed-graph | 0.5149 ± 0.0831 | 0.6714 ± 0.1106 |