Nonnegative Bayesian nonparametric factor models with completely random measures

Abstract We present a Bayesian nonparametric Poisson factorization model for modeling dense network data with an unknown and potentially growing number of overlapping communities. The construction is based on completely random measures allows the communities to either increase nodes at specified logarithmic or polynomial rate, be bounded. develop asymptotics size derive Markov chain Monte Carlo algorithm targeting exact posterior distribution this model. usefulness approach illustrated various real networks.