Node Classification On Non Homophilic 14

评估指标

1:1 Accuracy

评测结果

各个模型在此基准测试上的表现结果

		Paper Title
ClenshawGCN	91.69 ± 0.25	Clenshaw Graph Neural Networks
ACM-GCN	91.44 ± 0.08	Revisiting Heterophily For Graph Neural Networks
ACM-GCN++	91.4 ± 0.07	Revisiting Heterophily For Graph Neural Networks
ACM-GCN+	91.33 ± 0.11	Revisiting Heterophily For Graph Neural Networks
ACMII-GCN	91.19 ± 0.16	Revisiting Heterophily For Graph Neural Networks
ACMII-GCN+	91.13 ± 0.09	Revisiting Heterophily For Graph Neural Networks
ACMII-GCN++	91.01 ± 0.18	Revisiting Heterophily For Graph Neural Networks
GloGNN++	90.91 ± 0.13	Finding Global Homophily in Graph Neural Networks When Meeting Heterophily
OptBasisGNN	90.83±0.11	Graph Neural Networks with Learnable and Optimal Polynomial Bases
LINKX	90.77 ± 0.27	Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods
GloGNN	90.66 ± 0.11	Finding Global Homophily in Graph Neural Networks When Meeting Heterophily
MixHop	90.58 ± 0.16	MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing
GCNII	90.24 ± 0.09	Simple and Deep Graph Convolutional Networks
GPRGCN	90.05 ± 0.31	Adaptive Universal Generalized PageRank Graph Neural Network
GCNJK	89.30 ± 0.19	Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods
GCN	87.42 ± 0.37	Semi-Supervised Classification with Graph Convolutional Networks
MLP	86.68 ± 0.09	Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods
APPNP	85.36 ± 0.62	Predict then Propagate: Graph Neural Networks meet Personalized PageRank
C&S 2-hop	84.94 ± 0.49	Combining Label Propagation and Simple Models Out-performs Graph Neural Networks
C&S 1-hop	82.93 ± 0.15	Combining Label Propagation and Simple Models Out-performs Graph Neural Networks

0 of 26 row(s) selected.