PyG coding example (on node classification) for GCN: A choice of convolutional NN architecture motivated via a localized first-order approximation of spectral graph convolutions. The GCN model scales linearly in the number of graph edges and learns hidden layer representations that encode both local graph structure and features of nodes. SEMI-SUPERVISED CLASSIFICATION WITH GRAPH CONVOLUTIONAL NETWORKS Thomas N. Kipf, Max Welling https://arxiv.org/abs/1609.02907v4 All rights for code with PyG. See also: https://paperswithcode.com/method/gcn