Constraint Learning on Random constraints Dense synthetic (test)
96.6Vector AccuracyILP–Loss
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ILP–LossNumber of ground truth constraints=12022.10 | 96.6 | 7.3 | |
| CombOptNetNumber of ground truth constraints=12022.10 | 89.3 | 9.9 | |
| ILP–LossNumber of ground truth constraints=22022.10 | 86.3 | 15.6 | |
| CombOptNetNumber of ground truth constraints=22022.10 | 74.8 | 16.8 | |
| ILP–LossNumber of ground truth constraints=42022.10 | 74 | 17.6 | |
| ILP–LossNumber of ground truth constraints=82022.10 | 41.5 | 20.6 | |
| CombOptNetNumber of ground truth constraints=42022.10 | 34.3 | 24.7 | |
| CombOptNetNumber of ground truth constraints=82022.10 | 2 | 48.2 |