Random-Set Graph Neural Networks

2026-05-12

AI Summary

This work addresses the challenge of modeling node-level epistemic uncertainty in graph neural networks arising from uncertainties in graph structure or node features. It introduces belief function theory into graph neural network architectures for the first time, proposing a prediction head based on finite random sets that explicitly outputs class probabilities alongside a quantifiable measure of epistemic uncertainty. By integrating random set theory with graph neural networks, the method enables formal modeling and quantification of node-level epistemic uncertainty. Experimental evaluation across nine graph learning benchmarks—including autonomous driving datasets such as Nuscenes and ROAD—demonstrates that the proposed model significantly outperforms existing approaches in uncertainty quantification.

Abstract

Uncertainty quantification has become an important factor in understanding the data representations produced by Graph Neural Networks (GNNs). Despite their predictive capabilities being ever useful across industrial workspaces, the inherent uncertainty induced by the nature of the data is a huge mitigating factor to GNN performance. While aleatoric uncertainty is the result of noisy and incomplete stochastic data such as missing edges or over-smoothing, epistemic uncertainty arises from lack of knowledge about a system or model (e.g., a graph's topology or node feature representation), which can be reduced by gathering more data and information. In this paper, we propose an original new framework in which node-level epistemic uncertainty is modelled in a belief function (finite random set) formalism. The resulting Random-Set Graph Neural Networks have a belief-function head predicting a random set over the list of classes, from which both a precise probability prediction and a measure of epistemic uncertainty can be obtained. Extensive experiments on 9 different graph learning datasets, including real-world autonomous driving benchmarks as such Nuscene and ROAD, demonstrate RS-GNN's superior uncertainty quantification capabilities

Problem

• epistemic uncertainty
• Graph Neural Networks
• uncertainty quantification
• belief function
• random set

Innovation

• Random-Set GNN
• epistemic uncertainty
• belief function
• uncertainty quantification
• graph neural networks