Graph Classification On Nci109

Metrics

Accuracy

Results

Performance results of various models on this benchmark

Model Name	Accuracy	Paper Title	Repository
SAGPool_h	67.86	Self-Attention Graph Pooling
GIC	82.86	Gaussian-Induced Convolution for Graphs	-
Multigraph ChebNet	82.0	Spectral Multigraph Networks for Discovering and Fusing Relationships in Molecules
PNA	83.382±1.045	Principal Neighbourhood Aggregation for Graph Nets
WKPI-kcenters	87.3	Learning metrics for persistence-based summaries and applications for graph classification
GraphGPS	81.256±0.501	Recipe for a General, Powerful, Scalable Graph Transformer
Graph2Vec	74.26	graph2vec: Learning Distributed Representations of Graphs
CAN	83.6	Cell Attention Networks
GAT	82.560±0.601	Graph Attention Networks
DropGIN	83.961±1.141	DropGNN: Random Dropouts Increase the Expressiveness of Graph Neural Networks
GATv2	83.092±0.764	How Attentive are Graph Attention Networks?
HGP-SL	80.67	Hierarchical Graph Pooling with Structure Learning
S-CGIB	77.54±1.51	Pre-training Graph Neural Networks on Molecules by Using Subgraph-Conditioned Graph Information Bottleneck
GIUNet	77	Graph isomorphism UNet
PIN	84.0	Weisfeiler and Lehman Go Paths: Learning Topological Features via Path Complexes	-
Deep WL SGN(0,1,2)	71.06	Subgraph Networks with Application to Structural Feature Space Expansion	-
UGT	75.45±1.26	Transitivity-Preserving Graph Representation Learning for Bridging Local Connectivity and Role-based Similarity
Propagation kernels (pk)	83.5	Propagation kernels: efficient graph kernels from propagated information
GCN	83.140±1.248	Semi-Supervised Classification with Graph Convolutional Networks
ASAP	70.07	ASAP: Adaptive Structure Aware Pooling for Learning Hierarchical Graph Representations

0 of 37 row(s) selected.