Citeseer
Description
Citeseer
Dataset Overview
Lifting Methods
Structural-based Liftings
- Cellular: Cycle-based lifting
- Simplicial: Clique complex lifting
- Hypergraph: k-hop lifting
Feature Lifting
- Projected sum
Model Performance
| Model | Accuracy (%) | Std Dev (±) |
|---|---|---|
| GCN | 75.53 | 1.27 |
| GIN | 73.73 | 1.23 |
| GAT | 74.41 | 1.75 |
| AST | 73.85 | 2.21 |
| EDGNN | 74.93 | 1.39 |
| UniGNN2 | 74.72 | 1.08 |
| CWN | 75.20 | 1.82 |
| CCCN | 75.63 | 1.58 |
| SCCNN | 70.23 | 2.69 |
| SCN | 71.24 | 1.68 |
Key Insights
- CCCN achieves the best performance with 75.63% accuracy
- Most models perform in the 70-76% accuracy range
- AST shows the highest variability (±2.21)
- UniGNN2 shows the most stable results with lowest std dev (±1.08)
Reproducibility
python -m topobench model=cell/cccn dataset=graph/cocitation_citeseer optimizer.parameters.lr=0.001 model.feature_encoder.out_channels=128 model.backbone.n_layers=1 model.readout.readout_name=NoReadOut model.feature_encoder.proj_dropout=0.25 dataset.dataloader_params.batch_size=1 transforms.graph2cell_lifting.max_cell_length=10 dataset.split_params.data_seed=0,3,5,7,9 trainer.max_epochs=1000 trainer.min_epochs=50 trainer.check_val_every_n_epoch=1 callbacks.early_stopping.patience=50 logger.wandb.project=TopoBench --multirun