Convolutional Networks on Graphs for Learning Molecular Fingerprints

Benchmark Model Rank Results
drug-discovery-on-hiv-datasetGraphConv#2AUC: 0.822
drug-discovery-on-muvGraphConv#3AUC: 0.836
drug-discovery-on-toxcastGraphConv#3AUC: 0.754
drug-discovery-on-tox21GraphConv#5AUC: 0.846
node-classification-on-pubmed-0-03GCN-FP#10Accuracy: 56.2%
graph-regression-on-lipophilicityGC#11RMSE: 0.655
node-classification-on-cora-0-5GCN-FP#11Accuracy: 50.5%
node-classification-on-pubmed-0-05GCN-FP#11Accuracy: 63.2%
node-classification-on-pubmed-0-1GCN-FP#11Accuracy: 70.3%
node-classification-on-citeseer-0-5GCN-FP#12Accuracy: 43.9%
node-classification-on-citeseer-1GCN-FP#12Accuracy: 54.3%
node-classification-on-cora-1GCN-FP#12Accuracy: 59.6%
node-classification-on-cora-3GCN-FP#12Accuracy: 71.7%
node-classification-on-pubmed-with-public-split-fixed-20-nodes-per-classGCN-FP#36Accuracy: 76.0%
node-classification-on-cora-with-public-split-fixed-20-nodes-per-classGCN-FP#41Accuracy: 74.6%
node-classification-on-citeseer-with-public-split-fixed-20-nodes-per-classGCN-FP#45Accuracy: 61.5%