Hyperedge prediction on GPS non-k-uniform (test)
0.952AUCHyper-SAGNN-E
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Hyper-SAGNN-Etraining_data=hyperedges only2019.11 | 0.952 | 0.798 | |
| Hyper-SAGNN-E (mix)training_data=mixture of edges and hyper-edges2019.11 | 0.95 | 0.795 | |
| Hyper-SAGNN-Wtraining_data=hyperedges only2019.11 | 0.922 | 0.722 | |
| Hyper-SAGNN-W (mix)training_data=mixture of edges and hyper-edges2019.11 | 0.92 | 0.72 | |
| DHNE2019.11 | 0.91 | 0.668 | |
| node2vec minAggregation strategy=min2019.11 | 0.57 | 0.185 | |
| node2vec - meanAggregation strategy=mean2019.11 | 0.563 | 0.191 |