Node Classification On Citeseer

评估指标

Accuracy

Validation

评测结果

各个模型在此基准测试上的表现结果

			Paper Title	Repository
ACMII-Snowball-2	82.07 ± 1.04	-	Is Heterophily A Real Nightmare For Graph Neural Networks To Do Node Classification?	-
ACM-GCN	81.68 ± 0.97	-	Is Heterophily A Real Nightmare For Graph Neural Networks To Do Node Classification?	-
ACM-Snowball-2	81.58 ± 1.23	-	Is Heterophily A Real Nightmare For Graph Neural Networks To Do Node Classification?	-
ACMII-Snowball-3	81.56 ± 1.15	-	Is Heterophily A Real Nightmare For Graph Neural Networks To Do Node Classification?	-
SSP	80.52 ± 0.14	-	Optimization of Graph Neural Networks with Natural Gradient Descent
NodeNet	80.09%	-	NodeNet: A Graph Regularised Neural Network for Node Classification	-
SplineCNN	79.20%	-	SplineCNN: Fast Geometric Deep Learning with Continuous B-Spline Kernels
GCN-LPA	78.7 ± 0.6	-	Unifying Graph Convolutional Neural Networks and Label Propagation
Graph-MLP + SWA	77.99 ± 1.57%	-	The Split Matters: Flat Minima Methods for Improving the Performance of GNNs
CNMPGNN	76.81±1.40	-	CN-Motifs Perceptive Graph Neural Networks	-
CGT	76.59±0.98	-	Mitigating Degree Biases in Message Passing Mechanism by Utilizing Community Structures
3ference	76.33	-	Inferring from References with Differences for Semi-Supervised Node Classification on Graphs	-
MMA	76.30%	-	Multi-Mask Aggregators for Graph Neural Networks	-
AdaGCN	76.22 ± 0.20	-	AdaGCN: Adaboosting Graph Convolutional Networks into Deep Models
UGT	76.08±2.5	-	Transitivity-Preserving Graph Representation Learning for Bridging Local Connectivity and Role-based Similarity
PPNP	75.83%	YES	Predict then Propagate: Graph Neural Networks meet Personalized PageRank
CoLinkDist	75.79%	-	Distilling Self-Knowledge From Contrastive Links to Classify Graph Nodes Without Passing Messages
CoLinkDistMLP	75.77%	-	Distilling Self-Knowledge From Contrastive Links to Classify Graph Nodes Without Passing Messages
APPNP	75.73%	-	Predict then Propagate: Graph Neural Networks meet Personalized PageRank
Cleora	75.7	-	Cleora: A Simple, Strong and Scalable Graph Embedding Scheme

0 of 70 row(s) selected.