利用局部拓扑特征增强图表示学习
Enhancing Graph Representation Learning with Localized Topological Features
AI总结:
针对图神经网络表示能力有限的问题,提出基于持续同调理论提取图拓扑特征的方法,增强图表示学习性能,还探索了拓扑特征的端到端学习并提供理论与实证指导。
摘要:
图上的表示学习是一个基础问题,在各类任务中至关重要。图神经网络作为图表示学习的主流方法,其表示能力存在局限。因此,显式提取高阶拓扑与几何信息并融入这些模型会有所助益。本文基于持续同调理论,提出一种提取图丰富连通信息的系统性方法。该方法利用拓扑特征增强图神经网络的表示学习,在多个节点分类与链接预测基准上取得了最优性能。我们还探索了拓扑特征的端到端学习方案,即在学习过程中将拓扑计算视为可微算子。理论分析与实证研究为在图学习任务中使用拓扑特征提供了洞见与潜在指导。
英文摘要:
Representation learning on graphs is a fundamental problem that can be crucial in various tasks. Graph neural networks, the dominant approach for graph representation learning, are limited in their representation power. Therefore, it can be beneficial to explicitly extract and incorporate high-order topological and geometric information into these models. In this paper, we propose a principled approach to extract the rich connectivity information of graphs based on the theory of persistent homology. Our method utilizes the topological features to enhance the representation learning of graph neural networks and achieve state-of-the-art performance on various node classification and link prediction benchmarks. We also explore the option of end-to-end learning of the topological features, i.e., treating topological computation as a differentiable operator during learning. Our theoretical analysis and empirical study provide insights and potential guidelines for employing topological features in graph learning tasks.
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。