“…Tools that have been developed for such problems include, for example, graphon models (Borgs et al, 2008;Hoff et al, 2002;Bickel et al, 2011;Ambroise and Matias, 2012;Wolfe and Olhede, 2013;Gao et al, 2015), the subfamily of stochastic block models (e.g. Goldenberg et al (2010)), preferential attachment graphs (Barabási and Albert, 1999) and interacting particle models (Liggett, 2005). In this context, a statistical model is a family P = {P θ , θ ∈ T} of probability distributions on graphs, and sample data are explained either as a single graph drawn from the model, or as a subgraph of such a graph.…”