Graph Condensation via Receptive Field Distribution Matching

TL;DR

This paper introduces GCDM, a method for creating small, representative graphs by matching receptive field distributions, enabling efficient and generalizable GNN training.

Contribution

The paper proposes a novel graph condensation technique using receptive field distribution matching with MMD, improving speed and model generalization.

Findings

Synthetic graphs effectively match original receptive field distributions.

GCDM significantly accelerates the graph condensing process.

Condensed graphs generalize well across different GNN models.

Abstract

Graph neural networks (GNNs) enable the analysis of graphs using deep learning, with promising results in capturing structured information in graphs. This paper focuses on creating a small graph to represent the original graph, so that GNNs trained on the size-reduced graph can make accurate predictions. We view the original graph as a distribution of receptive fields and aim to synthesize a small graph whose receptive fields share a similar distribution. Thus, we propose Graph Condesation via Receptive Field Distribution Matching (GCDM), which is accomplished by optimizing the synthetic graph through the use of a distribution matching loss quantified by maximum mean discrepancy (MMD). Additionally, we demonstrate that the synthetic graph generated by GCDM is highly generalizable to a variety of models in evaluation phase and that the condensing speed is significantly improved using this framework.