Reviewed Version (pdf): https://openreview.net/references/pdf?id=zOb9dl2Q3q
Keywords: GAN, generative adversarial network, WGAN, GNN, graph neural network, generative model, graph
Abstract: We study the fundamental problem of graph generation. Specifically, we treat graph generation from a geometric perspective by associating each node with a position in space and then connecting the edges based on a similarity function. We then provide new solutions to the key challenges that prevent the widespread application of this classical geometric interpretation: (1) modeling complex relations, (2) modeling isomorphic graphs consistently, and (3) fully exploiting the latent distribution. Our main contribution is dubbed as the geometric graph (GG) generative adversarial network (GAN), which is a Wasserstein GAN that addresses the above challenges. GG-GAN is permutation equivariant and easily scales to generate graphs of tens of thousands of nodes. GG-GAN also strikes a good trade-off between novelty and modeling the distribution statistics, being competitive or surpassing the state-of-the-art methods that are either slower or that are non-equivariant, or that exploit problem-specific knowledge.
One-sentence Summary: We present the first fully equivariant pure GAN framework for graph generation, based on a geometric perspective on graphs and scalable for inference of thousands of nodes on commonly available hardware.
Supplementary Material: zip
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics