Graph Classification on PTC (Mean Accuracy and Stdev)
72.1Mean AccuracyMGNNI
Evaluation results
| Method | Links | ||
|---|---|---|---|
| MGNNIValidation=10-fold cross-validation, Setting=Inductive2022.10 | 72.1 | 2.8 | |
| IGNNValidation=10-fold cross-validation, Setting=Inductive2022.10 | 70.1 | 5.6 | |
| EIGNNValidation=10-fold cross-validation, Setting=Inductive2022.10 | 69.8 | 5.3 | |
| WKPIConfig Category=Top NT, Landmarks (K)=200, Evaluation Protocol=5 seeds x 10 folds2026.05 | 68.1 | — | |
| CGSValidation=10-fold cross-validation, Setting=Inductive2022.10 | 64.7 | 6.4 | |
| PLACELandmarks (K)=200, Evaluation Protocol=5 seeds x 10 folds2026.05 | 64.3 | — | |
| GCNValidation=10-fold cross-validation, Setting=Inductive2022.10 | 64.2 | 4.3 | |
| GINValidation=10-fold cross-validation, Setting=Inductive2022.10 | 63.7 | 8.2 | |
| FDGNNValidation=10-fold cross-validation, Setting=Inductive2022.10 | 63.4 | 5.4 | |
| PALACELandmarks (K)=200, Evaluation Protocol=5 seeds x 10 folds, Filtration Selection=closed-form gamma-hat/sqrt(K)-selected2026.05 | 63 | — | |
| DGCNNValidation=10-fold cross-validation, Setting=Inductive2022.10 | 58.6 | — |