arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2403.02723cs.AI
收藏

面向图神经网络的最小拓扑攻击

Minimum Topology Attacks for Graph Neural Networks

Beijing University of Posts and Telecommunications(北京邮电大学);Sony AI(索尼人工智能)

Mengmei Zhang, Xiao Wang, Chuan Shi, Lingjuan Lyu, Tianchi Yang, Junping Du

AI总结:

针对图神经网络固定预算拓扑攻击存在的预算适配困境,提出基于动态投影梯度下降的MiBTack最小预算拓扑攻击模型,可自适应找到最小攻击扰动,还能用于衡量节点鲁棒性并探究相关关系。

摘要:

随着图神经网络(Graph Neural Networks, GNNs)的广泛应用,其对抗拓扑攻击的鲁棒性受到了广泛关注。尽管已有诸多攻击方法被提出,但它们主要聚焦于固定预算攻击,旨在为目标节点找到固定预算内最具对抗性的扰动。然而,考虑到每个节点的鲁棒性各不相同,固定预算会带来不可避免的困境:当预算相对较小时,无法找到成功的扰动;而预算过大时,产生的冗余扰动又会损害攻击的隐蔽性。为打破这一困境,我们提出了一种新型拓扑攻击——最小预算拓扑攻击,旨在自适应地找到足以对每个节点实施成功攻击的最小扰动。为此,我们提出了名为MiBTack的攻击模型,该模型基于动态投影梯度下降算法,能够有效解决离散拓扑中涉及的非凸约束优化问题。在三种GNN和四个真实世界数据集上的大量实验结果表明,MiBTack能够以最少的扰动边数使所有目标节点被错误分类。此外,求得的最小预算可用于衡量节点鲁棒性,从而让我们能够探究节点的鲁棒性、拓扑结构与不确定性之间的关系,这是当前固定预算拓扑攻击无法实现的。

英文摘要:

With the great popularity of Graph Neural Networks (GNNs), their robustness to adversarial topology attacks has received significant attention. Although many attack methods have been proposed, they mainly focus on fixed-budget attacks, aiming at finding the most adversarial perturbations within a fixed budget for target node. However, considering the varied robustness of each node, there is an inevitable dilemma caused by the fixed budget, i.e., no successful perturbation is found when the budget is relatively small, while if it is too large, the yielding redundant perturbations will hurt the invisibility. To break this dilemma, we propose a new type of topology attack, named minimum-budget topology attack, aiming to adaptively find the minimum perturbation sufficient for a successful attack on each node. To this end, we propose an attack model, named MiBTack, based on a dynamic projected gradient descent algorithm, which can effectively solve the involving non-convex constraint optimization on discrete topology. Extensive results on three GNNs and four real-world datasets show that MiBTack can successfully lead all target nodes misclassified with the minimum perturbation edges. Moreover, the obtained minimum budget can be used to measure node robustness, so we can explore the relationships of robustness, topology, and uncertainty for nodes, which is beyond what the current fixed-budget topology attacks can offer.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑