Limits, approximation and size transferability for GNNs on sparse graphs via graphops
This work addresses the transferability challenge for GNNs in sparse graph domains, which is incremental as it extends existing dense graph theories to include sparse cases.
The authors tackled the problem of whether graph neural networks (GNNs) can generalize to graphs of different sizes, particularly sparse graphs like bounded-degree or power law graphs, by developing a theoretical framework using graphops. They derived quantitative bounds on the distance between finite GNNs and their limits on infinite graphs, as well as between GNNs on graphs of different sizes, applicable to both dense and sparse graphs.
Can graph neural networks generalize to graphs that are different from the graphs they were trained on, e.g., in size? In this work, we study this question from a theoretical perspective. While recent work established such transferability and approximation results via graph limits, e.g., via graphons, these only apply non-trivially to dense graphs. To include frequently encountered sparse graphs such as bounded-degree or power law graphs, we take a perspective of taking limits of operators derived from graphs, such as the aggregation operation that makes up GNNs. This leads to the recently introduced limit notion of graphops (Backhausz and Szegedy, 2022). We demonstrate how the operator perspective allows us to develop quantitative bounds on the distance between a finite GNN and its limit on an infinite graph, as well as the distance between the GNN on graphs of different sizes that share structural properties, under a regularity assumption verified for various graph sequences. Our results hold for dense and sparse graphs, and various notions of graph limits.