Graph Defense Diffusion Model

Xin He, Wenqi Fan, Yili Wang +4

Graph Neural Networks (GNNs) are highly vulnerable to adversarial attacks, which can greatly degrade their performance. Existing graph purification methods attempt to address this…