arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2310.19263cs.LG
收藏

一种元数据驱动的方法用于理解图神经网络

A Metadata-Driven Approach to Understand Graph Neural Networks

University of Michigan(密歇根大学);University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校)

Ting Wei Li, Qiaozhu Mei, Jiaqi Ma

AI总结:

本文提出元数据驱动方法,通过对GNN基准测试元数据进行多元稀疏回归识别关键数据属性,并以度分布为例通过理论与受控实验验证其对GNN性能的影响。

摘要:

图神经网络(GNNs)在各种应用中取得了显著成功,但其性能可能对其所处理的图数据集的特定数据属性敏感。当前理解GNN局限性的文献主要采用模型驱动的方法,利用网络科学或图论中的启发式和领域知识来建模GNN行为,这种方法耗时且高度主观。在这项工作中,我们提出一种元数据驱动的方法来分析GNN对图数据属性的敏感性,其动机是图学习基准的日益可用。我们对从跨不同数据集的GNN性能基准测试中得到的元数据进行多元稀疏回归分析,从而得到一组显著的数据属性。为验证我们数据驱动方法的有效性,我们聚焦于一个已识别的数据属性,即度分布,并通过理论分析和受控实验研究该属性如何影响GNN性能。我们的理论发现揭示,度分布更平衡的数据集表现出更好的节点表示线性可分性,从而带来更好的GNN性能。我们还使用具有不同度分布的合成数据集进行受控实验,结果与我们的理论发现高度一致。总体而言,理论分析和受控实验共同验证了所提出的元数据驱动方法在识别GNN关键数据属性方面是有效的。

英文摘要:

Graph Neural Networks (GNNs) have achieved remarkable success in various applications, but their performance can be sensitive to specific data properties of the graph datasets they operate on. Current literature on understanding the limitations of GNNs has primarily employed a $\textit{model-driven}$ approach that leverage heuristics and domain knowledge from network science or graph theory to model the GNN behaviors, which is time-consuming and highly subjective. In this work, we propose a $\textit{metadata-driven}$ approach to analyze the sensitivity of GNNs to graph data properties, motivated by the increasing availability of graph learning benchmarks. We perform a multivariate sparse regression analysis on the metadata derived from benchmarking GNN performance across diverse datasets, yielding a set of salient data properties. To validate the effectiveness of our data-driven approach, we focus on one identified data property, the degree distribution, and investigate how this property influences GNN performance through theoretical analysis and controlled experiments. Our theoretical findings reveal that datasets with more balanced degree distribution exhibit better linear separability of node representations, thus leading to better GNN performance. We also conduct controlled experiments using synthetic datasets with varying degree distributions, and the results align well with our theoretical findings. Collectively, both the theoretical analysis and controlled experiments verify that the proposed metadata-driven approach is effective in identifying critical data properties for GNNs.

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。

↑