this project implements graph convolutional networks with tensorflow 2.0. graph convolutional networks is a semi-supervised network which explicit represent the constraints with edges between nodes(samples). during training, the only partial labels need to be given for nodes(samples). the network will evolve to guess the label right for the unsupervised samples.
launch the training with
python3 train_cora.py