SIGN: Scalable Inception Graph Neural Networks

Benchmark Model Rank Results
node-classification-on-amz-compSIGN#5Accuracy: 85.93 ± 1.21
node-classification-on-amz-photoSIGN#13Accuracy: 91.72 ± 1.20
node-classification-on-coauthor-csSIGN#20Accuracy: 91.98 ± 0.50
node-classification-on-ppiSIGN#17F1: 96.50
node-classification-on-redditSIGN#7Accuracy: 96.60%
node-property-prediction-on-ogbn-arxivSIGN#63Test Accuracy: 0.7195 ± 0.0011Ext. data: No
node-property-prediction-on-ogbn-magSIGN#29Test Accuracy: 0.4046 ± 0.0012Ext. data: No
node-property-prediction-on-ogbn-papers100mSIGN-XL#13Test Accuracy: 0.6606 ± 0.0019Ext. data: No
node-property-prediction-on-ogbn-papers100mSIGN#15Test Accuracy: 0.6568 ± 0.0006Ext. data: No
node-property-prediction-on-ogbn-productsSIGN#39Test Accuracy: 0.8052 ± 0.0016Ext. data: No