Title: Stochastic Block Transition Models for Dynamic Networks

Abstract: There has been great interest in recent years in the development of statistical models for dynamic networks. This paper targets networks evolving in discrete time in which both nodes and edges can appear and disappear over time, such as dynamic networks of social interactions. We propose a stochastic block transition model (SBTM) for dynamic networks that is inspired by the well-known stochastic block model (SBM) for static networks and several recent dynamic extensions of the SBM. Unlike most existing dynamic models, it does not make a hidden Markov assumption on the edge-level dynamics, allowing the presence or absence of edges to directly influence future edge probabilities. We demonstrate that the proposed SBTM is significantly better at reproducing durations of edges in real social network data between edges while retaining the interpretability of the SBM.