AGNNCert: Defending Graph Neural Networks against Arbitrary Perturbations with Deterministic Certification

TL;DR

This paper introduces AGNNCert, a novel certified defense method that guarantees robustness of graph neural networks against any type of perturbation, including edges, nodes, and features, for node and graph classification tasks.

Contribution

AGNNCert is the first deterministic certified defense for GNNs against arbitrary perturbations, addressing limitations of existing methods and encompassing previous defenses as special cases.

Findings

Proves robustness guarantees against all perturbation types.

Outperforms state-of-the-art defenses on benchmark datasets.

Applicable to multiple GNN architectures and tasks.

Abstract

Graph neural networks (GNNs) achieve the state-of-the-art on graph-relevant tasks such as node and graph classification. However, recent works show GNNs are vulnerable to adversarial perturbations include the perturbation on edges, nodes, and node features, the three components forming a graph. Empirical defenses against such attacks are soon broken by adaptive ones. While certified defenses offer robustness guarantees, they face several limitations: 1) almost all restrict the adversary's capability to only one type of perturbation, which is impractical; 2) all are designed for a particular GNN task, which limits their applicability; and 3) the robustness guarantees of all methods except one are not 100% accurate. We address all these limitations by developing AGNNCert, the first certified defense for GNNs against arbitrary (edge, node, and node feature) perturbations with deterministic robustness guarantees, and applicable to the two most common node and graph classification tasks. AGNNCert also encompass existing certified defenses as special cases. Extensive evaluations on multiple benchmark node/graph classification datasets and two real-world graph datasets, and multiple GNNs validate the effectiveness of AGNNCert to provably defend against arbitrary perturbations. AGNNCert also shows its superiority over the state-of-the-art certified defenses against the individual edge perturbation and node perturbation.