The Logic of Graph Neural Networks

Graph neural networks (GNNs) are deep learning architectures for machine learning problems on graphs. It has recently been shown that the expressiveness of GNNs can be characterised precisely by the combinatorial Weisfeiler-Leman algorithms and by finite variable counting logics. The correspondence...

Full description

Saved in:
Bibliographic Details
Published inProceedings of the 36th Annual ACM/IEEE Symposium on Logic in Computer Science pp. 1 - 17
Main Author Grohe, Martin
Format Conference Proceeding
LanguageEnglish
Published IEEE 29.06.2021
Subjects
Computer architecture
Computer science
Deep learning
Graph neural networks
Machine learning algorithms
Online AccessGet full text
DOI10.1109/LICS52264.2021.9470677

Cover

More Information
Summary:Graph neural networks (GNNs) are deep learning architectures for machine learning problems on graphs. It has recently been shown that the expressiveness of GNNs can be characterised precisely by the combinatorial Weisfeiler-Leman algorithms and by finite variable counting logics. The correspondence has even led to new, higher-order GNNs corresponding to the WL algorithm in higher dimensions.The purpose of this paper is to explain these descriptive characterisations of GNNs.
DOI:10.1109/LICS52264.2021.9470677