Rethinking Graph Backdoor Attacks: A Distribution-Preserving Perspective

Graph Neural Networks (GNNs) have shown remarkable performance in various tasks. However, recent works reveal that GNNs are vulnerable to backdoor attacks. Generally, backdoor attack poisons the graph by attaching backdoor triggers and the target class label to a set of nodes in the training graph....

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	arXiv.org 2024-07
Hauptverfasser:	Zhang, Zhiwei, Lin, Minhua, Dai, Enyan, Wang, Suhang
Format:	Artikel
Sprache:	eng
Schlagworte:	Computer Science - Cryptography and Security Computer Science - Learning Data analysis Effectiveness Empirical analysis Graph neural networks Machine learning Nodes Outliers (statistics)
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!