Node Classification on Pubmed standard (test)
85.2AccuracyDFNet-ATT
Evaluation Results
| Method | Links | |
|---|---|---|
| DFNet-ATT2019.10 | 85.2 | |
| DFNet2019.10 | 84.3 | |
| DF-ATT2019.10 | 82.3 | |
| GATarchitecture=DenseBlock2019.10 | 81.8 | |
| DDMLearning Setup=unsupervised representation learning with linear classifier2023.06 | 81.7 | |
| GCNarchitecture=DenseBlock2019.10 | 81.5 | |
| GraphMix (GAT)Source=Our Experiments, Training Strategy=GraphMix, Backbone=GAT2019.09 | 81.1 | |
| GraphMAELearning Setup=unsupervised representation learning with linear classifier2023.06 | 81.1 | |
| CCA-SSGLearning Setup=unsupervised representation learning with linear classifier2023.06 | 81 | |
| GraphMix (GCN)Source=Our Experiments, Training Strategy=GraphMix, Backbone=GCN2019.09 | 80.98 | |
| DisenGCNSource=Reported in literature2019.09 | 80.5 | |
| GCN (self-training)Source=Our Experiments, Training Strategy=Self-training2019.09 | 80.42 | |
| MVGRLLearning Setup=unsupervised representation learning with linear classifier2023.06 | 80.1 | |
| BVATSource=Reported in literature2019.09 | 79.9 | |
| GCNSource=Our Experiments, Training Strategy=Supervised2019.09 | 79.86 | |
| GAT (self-training)Source=Our Experiments, Training Strategy=Self-training2019.09 | 79.67 | |
| Graph U-NetSource=Reported in literature2019.09 | 79.6 | |
| BGRLLearning Setup=unsupervised representation learning with linear classifier2023.06 | 79.6 | |
| InfoGCLLearning Setup=unsupervised representation learning with linear classifier2023.06 | 79.1 | |
| GATSource=Our Experiments, Training Strategy=Supervised2019.09 | 79.05 | |
| GCN2019.10 | 79 | |
| GAT2019.10 | 79 | |
| GCNSource=Reported in literature2019.09 | 79 | |
| GATSource=Reported in literature2019.09 | 79 | |
| GATLearning Setup=unsupervised representation learning with linear classifier2023.06 | 79 | |
| ARMA2019.10 | 78.9 | |
| MoNetSource=Reported in literature2019.09 | 78.8 | |
| GraphMix (Graph U-Net)Source=Our Experiments, Training Strategy=GraphMix, Backbone=Graph U-Net2019.09 | 78.76 | |
| LNet2019.10 | 78.3 | |
| Graph-U-Net (self-training)Source=Our Experiments, Training Strategy=Self-training2019.09 | 78.12 | |
| AdaLNet2019.10 | 78.1 | |
| Graph U-NetSource=Our Experiments, Training Strategy=Supervised2019.09 | 77.73 | |
| Balanced LL-orthogonal initializationLayers=402023.10 | 77.5 | |
| FCN (self-training)Source=Our Experiments, Training Strategy=Self-training2019.09 | 77.4 | |
| Balanced LL-orthogonal initializationLayers=202023.10 | 77.3 | |
| Planetoid2019.10 | 77.2 | |
| PlanetoidSource=Reported in literature2019.09 | 77.2 | |
| Balanced LL-orthogonal initializationLayers=102023.10 | 76.9 | |
| DGILearning Setup=unsupervised representation learning with linear classifier2023.06 | 76.8 | |
| GPT-GNNLearning Setup=unsupervised representation learning with linear classifier2023.06 | 76.3 | |
| Balanced LL-orthogonal initializationLayers=22023.10 | 76 | |
| Balanced LL-orthogonal initializationLayers=52023.10 | 75.7 | |
| Lipschitz normalizationLayers=2, Base Initialization=Xavier2023.10 | 74.8 | |
| Chebyshev2019.10 | 74.4 | |
| ChebyshevSource=Reported in literature2019.09 | 74.4 | |
| ICA2019.10 | 73.9 | |
| ICASource=Reported in literature2019.09 | 73.9 | |
| Lipschitz normalizationLayers=5, Base Initialization=Xavier2023.10 | 73.7 | |
| MT-GCNLabeled per class=7, Label rate=0.1%2020.09 | 73.1 | |
| Co-trainingLabeled per class=7, Label rate=0.1%2020.09 | 72.7 | |
| SemiEmbSource=Reported in literature2019.09 | 71.7 | |
| MLPSource=Reported in literature2019.09 | 71.4 | |
| SemiEmb2019.10 | 71.1 | |
| UnionLabeled per class=7, Label rate=0.1%2020.09 | 70.7 | |
| ManiRegSource=Reported in literature2019.09 | 70.7 | |
| M3SLabeled per class=7, Label rate=0.1%2020.09 | 70.6 | |
| MultiStageLabeled per class=7, Label rate=0.1%2020.09 | 70.2 | |
| MT-GCNLabeled per class=3, Label rate=0.05%2020.09 | 69.5 | |
| IntersectionLabeled per class=7, Label rate=0.1%2020.09 | 69.4 | |
| Co-trainingLabeled per class=3, Label rate=0.05%2020.09 | 68.3 | |
| Self-trainingLabeled per class=7, Label rate=0.1%2020.09 | 66.8 | |
| LPLabeled per class=3, Label rate=0.05%2020.09 | 66.4 | |
| GCN+VLabeled per class=7, Label rate=0.1%2020.09 | 65.9 | |
| MT-GCNLabeled per class=2, Label rate=0.03%2020.09 | 65.5 | |
| LPLabeled per class=7, Label rate=0.1%2020.09 | 65.4 | |
| DeepWalk2019.10 | 65.3 | |
| DeepWalkSource=Reported in literature2019.09 | 65.3 | |
| M3SLabeled per class=3, Label rate=0.05%2020.09 | 64.4 | |
| MultiStageLabeled per class=3, Label rate=0.05%2020.09 | 64.3 | |
| UnionLabeled per class=3, Label rate=0.05%2020.09 | 64 | |
| LP2019.10 | 63 | |
| LPSource=Reported in literature2019.09 | 63 | |
| Co-trainingLabeled per class=2, Label rate=0.03%2020.09 | 62.2 | |
| LPLabeled per class=2, Label rate=0.03%2020.09 | 61.4 | |
| GCN+VLabeled per class=2, Label rate=0.03%2020.09 | 60.5 | |
| IntersectionLabeled per class=3, Label rate=0.05%2020.09 | 59.3 | |
| M3SLabeled per class=2, Label rate=0.03%2020.09 | 59.2 | |
| Self-trainingLabeled per class=3, Label rate=0.05%2020.09 | 58.7 | |
| UnionLabeled per class=2, Label rate=0.03%2020.09 | 58.4 | |
| GCN+VLabeled per class=3, Label rate=0.05%2020.09 | 57.5 | |
| MultiStageLabeled per class=2, Label rate=0.03%2020.09 | 57.4 | |
| GCN-VLabeled per class=7, Label rate=0.1%2020.09 | 56.3 | |
| Lipschitz normalizationLayers=10, Base Initialization=Xavier2023.10 | 52.8 | |
| IntersectionLabeled per class=2, Label rate=0.03%2020.09 | 52 | |
| Self-trainingLabeled per class=2, Label rate=0.03%2020.09 | 51.9 | |
| ChebyshevLabeled per class=7, Label rate=0.1%2020.09 | 51.2 | |
| GCN-VLabeled per class=3, Label rate=0.05%2020.09 | 49.7 | |
| ChebyshevLabeled per class=3, Label rate=0.05%2020.09 | 47.3 | |
| GCN-VLabeled per class=2, Label rate=0.03%2020.09 | 46.4 | |
| ChebyshevLabeled per class=2, Label rate=0.03%2020.09 | 40.4 | |
| Lipschitz normalizationLayers=40, Base Initialization=Xavier2023.10 | 36.6 | |
| Lipschitz normalizationLayers=20, Base Initialization=Xavier2023.10 | 23.3 |