Graph Classification on DD
98.5AccuracyCDAT
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| CDATevaluation_protocol=100 runs of 10-fold cross-validation2026.06 | 98.5 | — | — | |
| ETevaluation_protocol=100 runs of 10-fold cross-validation2026.06 | 95.9 | — | — | |
| GraphMAEEvaluation protocol=Unsupervised representation learning2024.05 | 88.1 | — | — | |
| PXGL-GNN2025.12 | 86.54 | — | — | |
| S2GAE2025.12 | 84.3 | — | — | |
| SAGNN2025.12 | 84.12 | — | — | |
| ESA2024.02 | 83.5 | — | — | |
| ICL2025.12 | 82.77 | — | — | |
| SubGNN2025.12 | 82.51 | — | — | |
| MVGRLEvaluation protocol=Unsupervised representation learning2024.05 | 82.5 | — | — | |
| WKPI-KM2019.04 | 82 | — | — | |
| WKPIConfig Category=Top NT, Landmarks (K)=200, Evaluation Protocol=5 seeds x 10 folds2026.05 | 82 | — | — | |
| WKPI (k-means)evaluation_protocol=100 runs of 10-fold cross-validation2026.06 | 82 | — | — | |
| RetGK2019.04 | 81.6 | — | — | |
| RetGk2018.11 | 81.6 | — | — | |
| SAN2025.12 | 81.36 | — | — | |
| CTQWformer2026.05 | 81.24 | — | — | |
| CTQWformer2026.05 | 81.24 | — | — | |
| HGP-SLevaluation_protocol=100 runs of 10-fold cross-validation2026.06 | 81 | — | — | |
| CH-CL2026.04 | 80.92 | — | — | |
| GPS2024.02 | 80.8 | — | — | |
| Space OptimumRefinement budget=1002025.07 | 80.43 | — | — | |
| WKPI-KC2019.04 | 80.3 | — | — | |
| UGformerArchitecture=Variant 1, Inference Setting=Inductive, Evaluation Protocol=10-fold cross-validation2019.09 | 80.23 | — | — | |
| UGT# Parameters=76,9282023.11 | 80.23 | — | — | |
| UGT2023.11 | 80.23 | — | — | |
| M-DESIGNRefinement budget=1002025.07 | 80.14 | — | — | |
| BN-Pool2025.01 | 80 | — | — | |
| JTQK2026.05 | 79.89 | — | — | |
| RandomRefinement budget=1002025.07 | 79.86 | — | — | |
| WLSK2026.05 | 79.78 | — | — | |
| Auto-GNNRefinement budget=1002025.07 | 79.69 | — | — | |
| CI-GCLModel Type=Specialized graph-level GCL2025.11 | 79.63 | — | — | |
| CI-GCL2026.04 | 79.63 | — | — | |
| GT+GTE2026.03 | 79.6 | — | — | |
| GraphNASRefinement budget=1002025.07 | 79.57 | — | — | |
| PF2019.04 | 79.4 | — | — | |
| DGCNNInference Setting=Inductive, Evaluation Protocol=10-fold cross-validation2019.09 | 79.37 | — | — | |
| DGCNN2023.11 | 79.37 | — | — | |
| GFN# Parameters=68,7542023.11 | 79.37 | — | — | |
| DGCNN2023.11 | 79.37 | — | — | |
| GFN2023.11 | 79.37 | — | — | |
| DGCNN2026.05 | 79.37 | — | — | |
| GRAPHSAGE2025.12 | 79.24 | — | — | |
| RLRefinement budget=1002025.07 | 79.22 | — | — | |
| WL-OA2018.11 | 79.2 | — | — | |
| EARefinement budget=1002025.07 | 79.16 | — | — | |
| TopoGCL2026.04 | 79.15 | — | — | |
| GCN2026.05 | 79.12 | — | — | |
| GCN+2026.03 | 79.1 | — | — | |
| MGCLLearning Protocol=unsupervised representation learning, Evaluation Protocol=linear evaluation protocol, Cross-Validation=10-fold2025.10 | 79.07 | — | 1.42 | |
| DesiGNNRefinement budget=1002025.07 | 79.04 | — | — | |
| PNA2024.02 | 79 | — | — | |
| JBGNN2025.01 | 79 | — | — | |
| HOSC2025.01 | 79 | — | — | |
| CONG+backbone=GraphSage2025.12 | 78.9 | — | — | |
| RGCLLearning Protocol=unsupervised representation learning, Evaluation Protocol=linear evaluation protocol, Cross-Validation=10-fold2025.10 | 78.9 | — | 4.71 | |
| RGCL2026.04 | 78.86 | — | — | |
| GFNInference Setting=Inductive, Evaluation Protocol=10-fold cross-validation2019.09 | 78.78 | — | — | |
| GMCL2026.04 | 78.68 | — | — | |
| Graph-JEPAPre-training Type=Self-predictive2023.09 | 78.64 | — | — | |
| GraphCLModel Type=Specialized graph-level GCL2025.11 | 78.62 | — | — | |
| GraphCLmodel_category=graph-level models2025.09 | 78.62 | — | 4.67 | |
| GraphCLLearning Protocol=unsupervised representation learning, Evaluation Protocol=linear evaluation protocol, Cross-Validation=10-fold2025.10 | 78.62 | — | 6.42 | |
| GraphCL2026.04 | 78.62 | — | — | |
| WL2019.04 | 78.6 | — | — | |
| WL2018.11 | 78.6 | — | — | |
| GraphCLEvaluation protocol=Unsupervised representation learning2024.05 | 78.6 | — | — | |
| P-WL-UC2019.04 | 78.5 | — | — | |
| GK2018.11 | 78.5 | — | — | |
| CONG+backbone=GCN2025.12 | 78.5 | — | — | |
| CONG†backbone=GCN2025.12 | 78.5 | — | — | |
| GK2018.05 | 78.45 | — | — | |
| FD-MVGCL2025.11 | 78.45 | — | — | |
| GCGK2026.05 | 78.45 | — | — | |
| AutoTransferRefinement budget=1002025.07 | 78.44 | — | — | |
| Baseline2023.05 | 78.4 | — | — | |
| DRGCLModel Type=Specialized graph-level GCL2025.11 | 78.4 | — | — | |
| SimMLPadaptation=node-level GCL adapted to graph-level tasks2025.09 | 78.4 | — | 3.63 | |
| DRGCLmodel_category=graph-level models2025.09 | 78.4 | — | 2.83 | |
| DRGCLLearning Protocol=unsupervised representation learning, Evaluation Protocol=linear evaluation protocol, Cross-Validation=10-fold2025.10 | 78.4 | — | 3.57 | |
| GCN2024.02 | 78.2 | — | — | |
| DropGIN2024.02 | 78.2 | — | — | |
| LaGraphPre-training Type=Self-predictive2023.09 | 78.1 | — | — | |
| VGIB+GraphSAGEReadout function=Sum pooling, Backbone=GraphSAGE2021.12 | 78 | — | — | |
| MinCut2025.01 | 78 | — | — | |
| DMoN2025.01 | 78 | — | — | |
| WL2018.05 | 77.95 | — | — | |
| AD-GCL-FIXlabel_ratio=10%, backbone=GCN, protocol=10-Fold2021.06 | 77.91 | — | — | |
| NIDCLEvaluation protocol=Unsupervised representation learning2024.05 | 77.8 | — | — | |
| CONG†backbone=GraphSage2025.12 | 77.8 | — | — | |
| CONG+backbone=GAT2025.12 | 77.7 | — | — | |
| QJSK2026.05 | 77.68 | — | — | |
| GCAPSInference Setting=Inductive, Evaluation Protocol=10-fold cross-validation2019.09 | 77.62 | — | — | |
| CAPSGNN2026.05 | 77.62 | — | — | |
| RWGNN2026.05 | 77.6 | — | — | |
| AERK2026.05 | 77.6 | — | — | |
| AutoGCL2026.04 | 77.57 | — | — | |
| AutoGCLEvaluation protocol=Unsupervised representation learning2024.05 | 77.5 | — | — | |
| SimGRACEModel Type=Specialized graph-level GCL2025.11 | 77.44 | — | — |