发表机构
Hong Kong Baptist University; Renmin University of China; Chinese Academy of Sciences; Mohamed bin Zayed University of Artificial Intelligence; Sun Yat-sen University（香港浸会大学; 中国人民大学; 中国科学院; 穆罕默德·本·扎耶德人工智能大学; 中山大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究基于粗化的GNN训练在异构图上性能下降问题，提出自适应互补增强（ACE）策略，通过重新整合粗化丢弃信息、应用正则化及不确定性加权，在异构图基准上提升性能，同构图上保持竞争力且计算开销小。
AI 中文摘要
基于粗化的图神经网络（GNN）训练已成为将GNN扩展到大规模图的一个有前景的方向。然而，先前工作几乎只在同构图上评估，对更具挑战性的异构图设置探索不足。本文通过实验和理论表明，由于粗化过程中不可避免的图信息丢失，现有基于粗化的训练方法在异构图上性能显著下降。为解决此问题，提出了自适应互补增强（ACE），这是一种即插即用、与模型无关的策略，它重新整合粗化中丢弃的信息：ACE学习一个投影仪来重建原始节点特征，并应用各向异性结构正则化来嵌入局部异质性。还采用同方差不确定性加权来自适应平衡主要粗化图训练损失和全图辅助损失的组合训练目标。大量实验表明，ACE在异构图基准上带来一致的性能提升，同时在同构图上保持有竞争力的结果且计算开销最小。
英文摘要
Coarsening-based training for graph neural networks (GNNs), i.e.\ training on coarsened graphs rather than the original large ones, has become a promising direction for scaling GNNs to massive graphs. However, prior work has been evaluated almost exclusively on \textit{homophilic} graphs, leaving the more challenging \textit{heterophilic} settings underexplored. We show, both empirically and theoretically, that existing coarsening-based training methods suffer significant performance degradation on heterophilic graphs due to inevitable loss of graph information during coarsening. To address this, we propose {\bf A}daptive {\bf C}omplementary {\bf E}nhancement, a plug-and-play, model-agnostic strategy that reintegrates the information discarded in coarsening: ACE learns a projector for re-constructing original node features and applies \textit{anisotropic structural regularization} to embed local heterophily. We further adopt \textit{homoscedastic uncertainty weighting} to adaptively balance the combined training objective of primary coarsened-graph training loss and full-graph auxiliary loss with augmented node features re-constructed by the heterophily-aware projector. Extensive experiments show that ACE drives consistent gains on heterophilic benchmarks while preserving competitive results on homophilic graphs with minimal computational overhead. Code is available at the GitHub repository: https://github.com/vasile-paskardlgm/ACE.
CommentsAccepted at the 42nd Conference on Uncertainty in Artificial Intelligence, UAI 2026
