Navigating Complexity: Toward Lossless Graph Condensation via Expanding Window Matching

TL;DR

This paper introduces a novel lossless graph condensation method using expanding window matching and curriculum learning to better preserve original graph information, significantly improving the fidelity of condensed graphs for GNN training.

Contribution

It proposes the first lossless graph condensation approach by integrating diverse supervision signals through expanding window matching and curriculum learning.

Findings

Outperforms existing methods on multiple datasets.

Achieves near-perfect replication of original graph properties.

Enhances GNN training efficiency with smaller condensed graphs.

Abstract

Graph condensation aims to reduce the size of a large-scale graph dataset by synthesizing a compact counterpart without sacrificing the performance of Graph Neural Networks (GNNs) trained on it, which has shed light on reducing the computational cost for training GNNs. Nevertheless, existing methods often fall short of accurately replicating the original graph for certain datasets, thereby failing to achieve the objective of lossless condensation. To understand this phenomenon, we investigate the potential reasons and reveal that the previous state-of-the-art trajectory matching method provides biased and restricted supervision signals from the original graph when optimizing the condensed one. This significantly limits both the scale and efficacy of the condensed graph. In this paper, we make the first attempt toward \textit{lossless graph condensation} by bridging the previously neglected supervision signals. Specifically, we employ a curriculum learning strategy to train expert trajectories with more diverse supervision signals from the original graph, and then effectively transfer the information into the condensed graph with expanding window matching. Moreover, we design a loss function to further extract knowledge from the expert trajectories. Theoretical analysis justifies the design of our method and extensive experiments verify its superiority across different datasets. Code is released at https://github.com/NUS-HPC-AI-Lab/GEOM.