Node Classification On Pubmed 60 20 20 Random

평가 지표

1:1 Accuracy

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	1:1 Accuracy	Paper Title
ACM-Snowball-2	90.81 ± 0.52	Revisiting Heterophily For Graph Neural Networks
GCNII*	89.98 ± 0.52	Simple and Deep Graph Convolutional Networks
GCN+JK	90.09 ± 0.68	Revisiting Heterophily For Graph Neural Networks
ACMII-Snowball-3	91.31 ± 0.6	Revisiting Heterophily For Graph Neural Networks
ACM-SGC-1	87.75 ± 0.88	Revisiting Heterophily For Graph Neural Networks
MixHop	87.04 ± 4.10	MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing
NHGCN	91.56 ± 0.50	Neighborhood Homophily-Guided Graph Convolutional Network
ACM-GCN+	90.46 ± 0.69	Revisiting Heterophily For Graph Neural Networks
ACM-GCN++	90.39 ± 0.33	Revisiting Heterophily For Graph Neural Networks
GPRGNN	85.07 ± 0.09	Adaptive Universal Generalized PageRank Graph Neural Network
ACMII-GCN+	90.96 ± 0.62	Revisiting Heterophily For Graph Neural Networks
ACM-GCNII*	90.18 ± 0.51	Revisiting Heterophily For Graph Neural Networks
SGC-1	85.5 ± 0.76	Simplifying Graph Convolutional Networks
ACMII-GCN++	90.63 ± 0.56	Revisiting Heterophily For Graph Neural Networks
APPNP	85.02 ± 0.09	Predict then Propagate: Graph Neural Networks meet Personalized PageRank
ACMII-Snowball-2	90.56 ± 0.39	Revisiting Heterophily For Graph Neural Networks
H2GCN	87.78 ± 0.28	Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs
Snowball-3	88.8 ± 0.82	Break the Ceiling: Stronger Multi-scale Deep Graph Convolutional Networks
ACM-SGC-2	88.79 ± 0.5	Revisiting Heterophily For Graph Neural Networks
MLP-2	86.43 ± 0.13	Revisiting Heterophily For Graph Neural Networks

0 of 37 row(s) selected.