DERGMs: Degeneracy-restricted exponential random graph models

Abstract
We propose a new exponential family of models for random graphs. Starting from the standard exponential random graph model (ERGM) framework, we propose an extension that addresses some of the well-known issues with ERGMs. Specifically, we solve the problem of computational intractability and `degenerate' model behavior by an interpretable support restriction.
View on arXivComments on this paper