Abstract:Graph Neural Networks (GNN) are currently the most popular approach for learning and prediction on graph-structured data and are deployed in various fields, from social network analysis to drug discovery. However, there is limited mathematical understanding of the performance of GNNs. We discuss the various perspectives used to study statistical generalisation in GNNs. We identify three broad frameworks. The first approach, rooted in learning theory, relies on uniform convergence bounds and the complexity of the hypothesis class of specific GNN architectures. This approach also builds on the expressivity of GNNs, typically studied through the lens of graph isomorphism tests. The second principle is to simplify the neural architecture by analysing GNNs under the asymptotics of infinitely many parameters or infinite graph size. This approach approximates GNNs using Gaussian processes, neural tangent kernels or graphon neural network operators, which allow studying the generalisation or stability of trained GNNs. The third framework studies GNNs under random graph models, often the contextual stochastic block model, and derives non-asymptotic error rates using tools from high-dimensional statistics. We highlight some key theoretical results and discuss a few limitations and open research questions for each perspective.
| Comments: | 15 pages, 4 figures, submission for Special Issue in AStA Advances in Statistical Analysis |
| Subjects: | Methodology (stat.ME); Machine Learning (cs.LG); Machine Learning (stat.ML) |
| MSC classes: | 68Q32, 68T07, 62H12, 05C80 |
| Cite as: | arXiv:2605.25452 [stat.ME] |
| (or arXiv:2605.25452v1 [stat.ME] for this version) | |
| https://doi.org/10.48550/arXiv.2605.25452 arXiv-issued DOI via DataCite (pending registration) |
From: Debarghya Ghoshdastidar [view email]
[v1]
Mon, 25 May 2026 06:02:31 UTC (1,212 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。