1 paper
Yifan Qian, Paul Expert, Tom Rieu +2
We show that the classification performance of graph convolutional networks (GCNs) is related to the alignment between features, graph, and ground truth, which we quantify using a…