Maximilien Dreveton

We establish the theoretical limits of clustering under the Popularity-Adjusted Block Model and show that even without edge-density signals, cluster recovery is possible by leveraging differences in intra- and inter-cluster popularity parameters