Non-exchangeable Conformal Prediction for Temporal Graph Neural Networks

TL;DR

This paper introduces NCPNET, a novel conformal prediction framework for temporal graph neural networks that accounts for temporal dependencies, ensuring reliable uncertainty quantification in dynamic graph settings.

Contribution

It extends conformal prediction to dynamic graphs with a diffusion-based non-conformity score and an efficiency-aware optimization algorithm, addressing coverage violations caused by temporal dependencies.

Findings

Ensures guaranteed coverage in temporal graphs.

Reduces prediction set size by up to 31%.

Improves computational efficiency over existing methods.

Abstract

Conformal prediction for graph neural networks (GNNs) offers a promising framework for quantifying uncertainty, enhancing GNN reliability in high-stakes applications. However, existing methods predominantly focus on static graphs, neglecting the evolving nature of real-world graphs. Temporal dependencies in graph structure, node attributes, and ground truth labels violate the fundamental exchangeability assumption of standard conformal prediction methods, limiting their applicability. To address these challenges, in this paper, we introduce NCPNET, a novel end-to-end conformal prediction framework tailored for temporal graphs. Our approach extends conformal prediction to dynamic settings, mitigating statistical coverage violations induced by temporal dependencies. To achieve this, we propose a diffusion-based non-conformity score that captures both topological and temporal uncertainties within evolving networks. Additionally, we develop an efficiency-aware optimization algorithm that improves the conformal prediction process, enhancing computational efficiency and reducing coverage violations. Extensive experiments on diverse real-world temporal graphs, including WIKI, REDDIT, DBLP, and IBM Anti-Money Laundering dataset, demonstrate NCPNET's capability to ensure guaranteed coverage in temporal graphs, achieving up to a 31% reduction in prediction set size on the WIKI dataset, significantly improving efficiency compared to state-of-the-art methods. Our data and code are available at https://github.com/ODYSSEYWT/NCPNET.