A Network Science Perspective on Evaluating Deep Graph Generative Models
It is found that two deep graph generative models produce synthetic networks that closely resemble the structural properties of real-world networks, enabling them to identify effective immunization strategies.