Undirected network models with degree heterogeneity and homophily

摘要: The degree heterogeneity and homophily are two typical features in network data. In this paper, we formulate a general model for undirected networks with these present the moment estimation inferring parameters. Our only specifies marginal distribution on each edge weighted or unweighted graphs admits non-independent dyad structures unlike previous works that assume independent dyads. We establish unified theoretical framework under which consistency of estimator hold as size goes to infinity. also derive its asymptotic representation can be used characterize limiting distribution. parameter contains bias term. Accurate inference necessitates bias-correction.Several applications provided illustrate result.