arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2404.04559cs.LGcs.NAeess.SPmath.NA
收藏

通过二维(2-D)图卷积实现的谱图神经网络

Spectral GNN via Two-dimensional (2-D) Graph Convolution

MBZUAI(穆罕默德·本·扎耶德人工智能大学);Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所);Florida International University(佛罗里达国际大学)

Guoming Li, Jian Yang, Shangsong Liang, Dongsheng Luo

AI总结:

针对现有谱 GNN 卷积无法构造目标输出的缺陷,本文提出统一各类范式的 2-D 图卷积及 ChebNet2D 实现,并验证其有效性与效率。

摘要:

谱图神经网络(GNN)已在图学习中取得巨大成功。作为谱 GNN 的关键组成部分,谱图卷积提取图数据中的重要频率信息,使谱 GNN 在下游任务中表现优异。然而,本文指出现有谱 GNN 在执行谱图卷积时仍存在关键缺陷。具体而言,将谱图卷积视为构造目标输出的操作时,我们证明,在输入图信号满足温和条件的情况下,现有主流卷积范式无法构造目标输出,从而导致谱 GNN 陷入次优解。为解决这些问题,我们从更一般的二维(2-D)信号卷积视角重新思考谱图卷积,并提出一种名为 2-D 图卷积的新卷积范式。我们证明,2-D 图卷积统一了现有图卷积范式,并能够构造任意目标输出。基于所提出的 2-D 图卷积,我们进一步提出 ChebNet2D,它通过应用 Chebyshev 插值,对 2-D 图卷积进行高效且有效的 GNN 实现。在基准数据集上的大量实验证明了 ChebNet2D 的有效性和效率。

英文摘要:

Spectral Graph Neural Networks (GNNs) have achieved tremendous success in graph learning. As an essential part of spectral GNNs, spectral graph convolution extracts crucial frequency information in graph data, leading to superior performance of spectral GNNs in downstream tasks. However, in this paper, we show that existing spectral GNNs remain critical drawbacks in performing the spectral graph convolution. Specifically, considering the spectral graph convolution as a construction operation towards target output, we prove that existing popular convolution paradigms cannot construct the target output with mild conditions on input graph signals, causing spectral GNNs to fall into suboptimal solutions. To address the issues, we rethink the spectral graph convolution from a more general two-dimensional (2-D) signal convolution perspective and propose a new convolution paradigm, named 2-D graph convolution. We prove that 2-D graph convolution unifies existing graph convolution paradigms, and is capable to construct arbitrary target output. Based on the proposed 2-D graph convolution, we further propose ChebNet2D, an efficient and effective GNN implementation of 2-D graph convolution through applying Chebyshev interpolation. Extensive experiments on benchmark datasets demonstrate both effectiveness and efficiency of the ChebNet2D.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑