Skip to yearly menu bar Skip to main content


Poster

OOD Link Prediction Generalization Capabilities of Message-Passing GNNs in Larger Test Graphs

Yangze Zhou · Gitta Kutyniok · Bruno Ribeiro

Keywords: [ graphon ] [ link prediction ] [ OOD ] [ Message Passing GNNs ] [ random graphs ] [ GNNs ]


Abstract:

This work provides the first theoretical study on the ability of graph Message Passing Neural Networks (gMPNNs) ---such as Graph Neural Networks (GNNs)--- to perform inductive out-of-distribution (OOD) link prediction tasks, where deployment (test) graph sizes are larger than training graphs. We first prove non-asymptotic bounds showing that link predictors based on permutation-equivariant (structural) node embeddings obtained by gMPNNs can converge to a random guess as test graphs get larger. We then propose a theoretically-sound gMPNN that outputs structural pairwise (2-node) embeddings and prove non-asymptotic bounds showing that, as test graphs grow, these embeddings converge to embeddings of a continuous function that retains its ability to predict links OOD. Empirical results on random graphs show agreement with our theoretical results.

Chat is not available.