Detectability thresholds and optimal algorithms for community structure in dynamic networks
This work addresses the fundamental limits of community detection in dynamic networks, providing theoretical and algorithmic insights for researchers in network science and machine learning, though it is incremental as it builds on existing dynamic stochastic block models.
The paper tackles the problem of learning latent community structure in dynamic networks with changing node memberships, deriving the exact detectability threshold below which no algorithm can perform better than chance and presenting two optimal algorithms that succeed down to this limit.
We study the fundamental limits on learning latent community structure in dynamic networks. Specifically, we study dynamic stochastic block models where nodes change their community membership over time, but where edges are generated independently at each time step. In this setting (which is a special case of several existing models), we are able to derive the detectability threshold exactly, as a function of the rate of change and the strength of the communities. Below this threshold, we claim that no algorithm can identify the communities better than chance. We then give two algorithms that are optimal in the sense that they succeed all the way down to this limit. The first uses belief propagation (BP), which gives asymptotically optimal accuracy, and the second is a fast spectral clustering algorithm, based on linearizing the BP equations. We verify our analytic and algorithmic results via numerical simulation, and close with a brief discussion of extensions and open questions.