GCN Related Work. Inspired from https://github.com/tkipf/gcn
- Monoplex Network (a.k.a Singal Network)
- Multiplex Network (Each layer has exact same number of nodes)
- Multilayer Network (Each Layer does not need to have same number of nodes)
- Input adj shape [3D Tensor]:
shape = [num_layer, num_nodes, num_nodes]
- Input Features shape [3D Tensor]:
shape = [num_layer, num_nodes, num_features]