Traffic Forecasting on Traffic4cast Berlin (test)
0.0667MAERCSNet
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| RCSNetMode=Same-city training/testing2026.05 | 0.0667 | 0.4089 | 0.6395 | |
| Multi-task Transfer LearningMode=Same-city training/testing2026.05 | 0.0745 | 0.4512 | 0.6717 | |
| UNetMode=Same-city training/testing2026.05 | 0.0782 | 0.4689 | 0.6848 | |
| U-Net + TDAMode=Same-city training/testing2026.05 | 0.0856 | 0.5234 | 0.7235 | |
| SwinUnet3DMode=Same-city training/testing2026.05 | 0.0867 | 0.5289 | 0.7273 | |
| 3DResNet + Sparse-UNetMode=Same-city training/testing2026.05 | 0.0889 | 0.5401 | 0.7349 | |
| Traffic Movie SnippetsMode=Same-city training/testing2026.05 | 0.0912 | 0.5567 | 0.7461 | |
| Graph-based U-NetMode=Same-city training/testing2026.05 | 0.0934 | 0.5678 | 0.7535 | |
| Dual Encoding U-NetMode=Same-city training/testing2026.05 | 0.0956 | 0.5789 | 0.7609 | |
| Historical AverageMode=Same-city training/testing2026.05 | 0.2324 | 1.1583 | 1.0762 |