Learning Graph Neural Networks with Approximate Gradient Descent

The dataset used in the paper is a graph neural network (GNN) dataset, where the goal is to learn a GNN with one hidden layer for node information convolution.

BibTex: