Inference for a generalised stochastic block model with unknown number\n of blocks and non-conjugate edge models

2019/09/20 by Matthew Ludkin, Ludkin, Matthew
Physics and Astronomy · #Complex Network Analysis Techniques #FOS: Computer and information sciences #Methodology (stat.ME) #Opinion Dynamics and Social Influence

paper · pdf · doi:10.48550/arxiv.1909.09421

openalex publication_date 2019/09/20 · openalex created_date 2022/07/28 · openalex updated_date 2026/07/28

Abstract

The stochastic block model (SBM) is a popular model for capturing community\nstructure and interaction within a network. Network data with non-Boolean edge\nweights is becoming commonplace; however, existing analysis methods convert\nsuch data to a binary representation to apply the SBM, leading to a loss of\ninformation. A generalisation of the SBM is considered, which allows edge\nweights to be modelled in their recorded state. An effective reversible jump\nMarkov chain Monte Carlo sampler is proposed for estimating the parameters and\nthe number of blocks for this generalised SBM. The methodology permits\nnon-conjugate distributions for edge weights, which enable more flexible\nmodelling than current methods as illustrated on synthetic data, a network of\nbrain activity and an email communication network.\n

Related

• Adapting the Stochastic Block Model to Edge-Weighted Networks
• Estimating Stochastic Block Models in the Presence of Covariates
• Variable selection for (realistic) stochastic blockmodels
• Blockmodels: A R-package for estimating in Latent Block Model and\n Stochastic Block Model, with various probability functions, with or without\n covariates
• Exact integrated completed likelihood maximisation in a stochastic block transition model for dynamic networks