Pruning Spurious Subgraphs for Graph Out-of-Distribtuion Generalization

2025-06-06

AI Summary

Graph Neural Networks (GNNs) suffer degraded out-of-distribution (OOD) generalization under distribution shifts due to spurious edges that introduce non-causal correlations. To address this, we propose PrunE—the first structure-pruning-based graph OOD method—which abandons the error-prone paradigm of directly identifying invariant subgraphs and instead actively prunes spurious edges strongly correlated with labels yet causally irrelevant, thereby preserving causal invariant subgraphs more completely. PrunE introduces two novel regularizers: a graph-size constraint and ε-probabilistic alignment, jointly enforcing structural sparsity and distributional consistency across environments. It further integrates distributionally robust optimization with theoretically guaranteed invariance. Extensive experiments on multiple graph OOD benchmarks demonstrate significant improvements over state-of-the-art methods. We provide theoretical analysis proving that the pruned model achieves stronger preservation of invariant features and enhanced OOD robustness.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Reasoning under Uncertainty: Graphical Models
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Graph Neural Networks (GNNs) often encounter significant performance degradation under distribution shifts between training and test data, hindering their applicability in real-world scenarios. Recent studies have proposed various methods to address the out-of-distribution generalization challenge, with many methods in the graph domain focusing on directly identifying an invariant subgraph that is predictive of the target label. However, we argue that identifying the edges from the invariant subgraph directly is challenging and error-prone, especially when some spurious edges exhibit strong correlations with the targets. In this paper, we propose PrunE, the first pruning-based graph OOD method that eliminates spurious edges to improve OOD generalizability. By pruning spurious edges, mine{} retains the invariant subgraph more comprehensively, which is critical for OOD generalization. Specifically, PrunE employs two regularization terms to prune spurious edges: 1) graph size constraint to exclude uninformative spurious edges, and 2) $epsilon$-probability alignment to further suppress the occurrence of spurious edges. Through theoretical analysis and extensive experiments, we show that PrunE achieves superior OOD performance and outperforms previous state-of-the-art methods significantly. Codes are available at: href{https://github.com/tianyao-aka/PrunE-GraphOOD}{https://github.com/tianyao-aka/PrunE-GraphOOD}.

Problem

• Improves GNN generalization by pruning spurious edges
• Addresses distribution shifts between training and test data
• Identifies invariant subgraphs for better OOD performance

Innovation

• Prunes spurious edges for OOD generalization
• Uses graph size constraint regularization
• Employs ε-probability alignment regularization