Node-level Regression Uncertainty Quantification on Twitch
100PICPBayesianNN
Evaluation results
| Method | Links | ||
|---|---|---|---|
| BayesianNN2026.05 | 100 | 3.07 | |
| ER-GNN2026.05 | 99 | 1.33 | |
| QpiGNNWidth penalty (lambda)=0.12026.05 | 98 | 0.54 | |
| QpiGNNWidth penalty (lambda)=opt.2026.05 | 94 | 0.36 | |
| CF-GNN2026.05 | 92 | 3.53 | |
| RQRadj.-GNN2026.05 | 91 | 0.42 | |
| MC Dropout2026.05 | 91 | 0.15 | |
| CF-GNNoptimized=true2026.05 | 89 | 2.34 | |
| QpiGNNWidth penalty (lambda)=0.52026.05 | 59 | 0.08 | |
| SQR-GNN2026.05 | 30 | 0.03 |