Pseudo-likelihood methods for community detection in large sparse networks

TL;DR

This paper introduces a fast pseudo-likelihood approach for community detection in large, sparse networks, improving scalability and accuracy over existing methods, and includes spectral clustering with perturbations for initialization.

Contribution

The paper presents a novel pseudo-likelihood algorithm for stochastic block models, including a degree-conditioned variant, and introduces spectral clustering with perturbations for sparse networks.

Findings

Algorithms perform well on very sparse networks

Pseudo-likelihood provides consistent community estimates

Spectral clustering with perturbations improves initialization

Abstract

Many algorithms have been proposed for fitting network models with communities, but most of them do not scale well to large networks, and often fail on sparse networks. Here we propose a new fast pseudo-likelihood method for fitting the stochastic block model for networks, as well as a variant that allows for an arbitrary degree distribution by conditioning on degrees. We show that the algorithms perform well under a range of settings, including on very sparse networks, and illustrate on the example of a network of political blogs. We also propose spectral clustering with perturbations, a method of independent interest, which works well on sparse networks where regular spectral clustering fails, and use it to provide an initial value for pseudo-likelihood. We prove that pseudo-likelihood provides consistent estimates of the communities under a mild condition on the starting value, for the case of a block model with two communities.