Learning Graph While Training: An Evolving Graph Convolutional Neural Network

Ruoyu Li, Junzhou Huang

Convolution Neural Networks on Graphs are important generalization and extension of classical CNNs. While previous works generally assumed that the graph structures of samples are…