NCI109
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Graph Classification | NCI109 | Accuracy89.3 | 267 | |
| Graph Classification | NCI109 (test) | Accuracy87.4 | 64 | |
| Graph Classification | NCI109 (10-fold cross-validation) | Accuracy87.4 | 32 | |
| Graph Classification | NCI109 TUDataset | Accuracy90.5 | 30 | |
| Graph-level molecular classification | NCI109 | MCC0.7 | 18 | |
| Graph Classification | NCI109 size shift (test) | MCC0.37 | 17 | |
| Backdoor Attack | NCI109 | Attack Success Rate (ASR)59 | 15 | |
| Graph similarity ranking | NCI109 | Kendall's Tau0.673 | 14 | |
| Graph Classification | NCI109 TU (unsupervised) | Accuracy79.7 | 11 | |
| Graph-level anomaly detection | NCI109 | AUROC76.12 | 7 | |
| Graph Classification | NCI109 Fair setup | Accuracy85.4 | 6 | |
| Graph Classification | NCI109 TU (test) | Accuracy85.7 | 6 | |
| Backdoor Attack on Federated Graph Learning | NCI109 | AAS64 | 5 | |
| Adversarial Attack | NCI109 | Average Attack Success (AAS)65 | 5 | |
| Graph Backdoor Attack | NCI109 | AAS64 | 5 |
Showing 15 of 15 rows