A network Poisson model for weighted directed networks with covariates

The edges in networks are not only binary, either present or absent, but also take weighted values in many scenarios (e.g., the number of emails between two users). The covariate-p(0) model has been proposed to model binary directed networks with the degree heterogeneity and covariates. However, it may cause information loss when it is applied in weighted networks. In this paper, we propose to use the Poisson distribution to model weighted directed networks, which admits the sparsity of networks, the degree heterogeneity and the homophily caused by covariates of nodes. We call it the network Poisson model. The model contains a density parameter mu, a 2n-dimensional node parameter theta and a fixed dimensional regression coefficient gamma of covariates. Since the number of parameters increases with n, asymptotic theory is non standard. When the number n of nodes goes to infinity, we establish the l(infinity)-errors for the maximum likelihood estimators (MLEs), (theta) over cap and (gamma) over cap, which are O-p((log n/n)(1/2)) for theta and O-p(log n/n) for (gamma) over cap, up to an additional factor. We also obtain the asymptotic normality of the MLE. Numerical studies and a data analysis demonstrate our theoretical findings.

Automated summary

The proposed network Poisson model is used for simulating weighted directed networks, taking into account the sparsity, degree heterogeneity, and homophily caused by node covariates. The research shows that, as the number of nodes approaches infinity, the maximum likelihood estimators can achieve a certain level of accuracy.