Extended stochastic block models with application to criminal networks

DOI10.1214/21-AOAS1595MaRDI QIDQ2080752zbMATH OpenOpenAlexFDO

Authors Sirio Legramanti, Tommaso Rigon, Daniele Durante, David Dunson

Publication date 10 October 2022

Published in The Annals of Applied Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2007.08569

zbMATH Keywords

networkBayesian nonparametricsproduct partition modelGibbs-type prior

Mathematics Subject Classification ID

Bayesian inference (62F15) Nonparametric estimation (62G05) Classification and discrimination; cluster analysis (statistical aspects) (62H30) Applications of statistics to social sciences (62P25) Social networks; opinion dynamics (91D30)

Abstract: Reliably learning group structures among nodes in network data is challenging in several applications. We are particularly motivated by studying covert networks that encode relationships among criminals. These data are subject to measurement errors, and exhibit a complex combination of an unknown number of core-periphery, assortative and disassortative structures that may unveil key architectures of the criminal organization. The coexistence of these noisy block patterns limits the reliability of routinely-used community detection algorithms, and requires extensions of model-based solutions to realistically characterize the node partition process, incorporate information from node attributes, and provide improved strategies for estimation and uncertainty quantification. To cover these gaps, we develop a new class of extended stochastic block models (ESBM) that infer groups of nodes having common connectivity patterns via Gibbs-type priors on the partition process. This choice encompasses many realistic priors for criminal networks, covering solutions with fixed, random and infinite number of possible groups, and facilitates the inclusion of node attributes in a principled manner. Among the new alternatives in our class, we focus on the Gnedin process as a realistic prior that allows the number of groups to be finite, random and subject to a reinforcement process coherent with criminal networks. A collapsed Gibbs sampler is proposed for the whole ESBM class, and refined strategies for estimation, prediction, uncertainty quantification and model selection are outlined. The ESBM performance is illustrated in realistic simulations and in an application to an Italian mafia network, where we unveil key complex block structures, mostly hidden from state-of-the-art alternatives.

Recommendations

• Inference for a generalised stochastic block model with unknown number of blocks and non-conjugate edge models
• A BAYESIAN APPROACH TO MODELING STOCHASTIC BLOCKSTRUCTURES WITH COVARIATES
• Dynamic stochastic block models: parameter estimation and detection of changes in community structure
• Stochastic blockmodels with a growing number of classes
• Classification and estimation in the stochastic blockmodel based on the empirical degrees

Cites work

• scientific article; zbMATH DE number 6114089 (Why is no real title available?)
• scientific article; zbMATH DE number 2214054 (Why is no real title available?)
• A BAYESIAN APPROACH TO MODELING STOCHASTIC BLOCKSTRUCTURES WITH COVARIATES
• A Block Model for Node Popularity in Networks with Community Structure
• A Consistent Adjacency Spectral Embedding for Stochastic Blockmodel Graphs
• A mixture of experts latent position cluster model for social network data
• A new Monte Carlo method for estimating marginal likelihoods
• A species sampling model with finitely many types
• A widely applicable Bayesian information criterion
• An asymptotic analysis of a class of discrete nonparametric priors
• Analysis of spectral clustering algorithms for community detection: the general bipartite setting
• Asymptotic equivalence of Bayes cross validation and widely applicable information criterion in singular learning theory
• Asymptotic normality of maximum likelihood and its variational approximation for stochastic blockmodels
• Bayes Factors
• Bayesian Clustering and Product Partition Models
• Bayesian Measures of Model Complexity and Fit
• Bayesian Nonparametric Estimation of the Probability of Discovering New Species
• Bayesian cluster analysis: point estimation and credible balls (with discussion)
• Bayesian generalized product partition model
• Bayesian nonparametric estimators derived from conditional Gibbs structures
• Bayesian testing for exogenous partition structures in stochastic block models
• Community detection and stochastic block models: recent developments
• Community detection in networks with node features
• Community structure in social and biological networks
• Comparing clusterings -- an information based distance
• Consistency of community detection in networks under degree-corrected stochastic block models
• Consistency of spectral clustering in stochastic block models
• Controlling the Reinforcement in Bayesian Non-Parametric Mixture Models
• Covariate-assisted spectral clustering
• Estimating the marginal likelihood using the arithmetic mean identity
• Estimating the number of communities by spectral methods
• Estimation and Prediction for Stochastic Blockstructures
• Fast unfolding of communities in large networks
• Global spectral clustering in dynamic networks
• Hierarchical clustering with discrete latent variable models and the integrated classification likelihood
• Identifying overlapping terrorist cells from the Noordin Top actor-event network
• Inconsistency of Pitman-Yor process mixtures for the number of components
• Likelihood-based model selection for stochastic block models
• Mixed membership stochastic blockmodels
• Mixture models with a prior on the number of components
• Model selection and clustering in stochastic block models based on the exact integrated complete data likelihood
• Multiresolution Network Models
• Network cross-validation by edge sampling
• Network cross-validation for determining the number of communities in network data
• Probabilistic Community Detection With Unknown Number of Communities
• Pseudo-likelihood methods for community detection in large sparse networks
• Role of normalization in spectral clustering for stochastic blockmodels
• Spectral clustering and the high-dimensional stochastic blockmodel
• Statistical inference on random dot product graphs: a survey
• The hierarchy of block models
• Understanding predictive information criteria for Bayesian models

Cited in

(6)

• Finite-dimensional Discrete Random Structures and Bayesian Clustering
• Dynamic Stochastic Blockmodel Regression for Network Data: Application to International Militarized Conflicts
• Extended stochastic block model with spatial covariates for weighted brain networks
• Horizon-\(K\) farsightedness in criminal networks
• Bayesian mixture models (in)consistency for the number of clusters
• Bayesian learning of graph substructures

This page was built for publication: Extended stochastic block models with application to criminal networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2080752)