Graph Condensation via Receptive Field Distribution Matching
This addresses the challenge of efficient graph representation for GNN training, which is incremental as it builds on existing graph condensation methods by introducing a novel distribution matching approach.
The paper tackles the problem of creating a small synthetic graph that accurately represents a larger original graph for training graph neural networks (GNNs), achieving this by matching receptive field distributions using maximum mean discrepancy (MMD) to optimize the synthetic graph, with results showing high generalizability across models and significantly improved condensing speed.
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.