Blockmodels: A R-package for estimating in Latent Block Model and Stochastic Block Model, with various probability functions, with or without covariates

TL;DR

This paper introduces an R-package implementing Variational EM algorithms for Stochastic and Latent Block Models, enabling efficient analysis of large networks with various probability functions and covariates, including automatic model selection.

Contribution

The package provides a flexible, efficient implementation of blockmodel estimation with covariates, supporting multiple probability functions and automatic group number selection.

Findings

Supports large networks with thousands of nodes

Allows analysis with covariates and multiple probability functions

Enables automatic model selection via ICL criterion

Abstract

Analysis of the topology of a graph, regular or bipartite one, can be done by clustering for regular ones or co-clustering for bipartite ones. The Stochastic Block Model and the Latent Block Model are two models, which are very similar for respectively regular and bipartite graphs, based on probabilistic models. Initially developed for binary graphs, these models have been extended to valued networks with optional covariates on the edges. This paper present a implementation of a Variational EM algorithm for Stochastic Block Model and Latent Block Model for some common probability functions, Bernoulli, Gaussian and Poisson, without or with covariates, with some standard flavors, like multivariate extensions. This implementation allow automatic group number exploration and selection via the ICL criterion, and allow analyze networks with thousands of nodes in a reasonable amount of time.