Likelihood-based model selection for stochastic block models

DOI10.1214/16-AOS1457zbMATH Open1371.62017arXiv1502.02069OpenAlexW2963873688WikidataQ105584288 ScholiaQ105584288MaRDI QIDQ90065FDOQ90065

Y. X. Rachel Wang, Peter J. Bickel, Y. X. Rachel Wang, Yanyan Li

Publication date: 1 April 2017

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

Abstract: The stochastic block model (SBM) provides a popular framework for modeling community structures in networks. However, more attention has been devoted to problems concerning estimating the latent node labels and the model parameters than the issue of choosing the number of blocks. We consider an approach based on the log likelihood ratio statistic and analyze its asymptotic properties under model misspecification. We show the limiting distribution of the statistic in the case of underfitting is normal and obtain its convergence rate in the case of overfitting. These conclusions remain valid when the average degree grows at a polylog rate. The results enable us to derive the correct order of the penalty term for model complexity and arrive at a likelihood-based model selection criterion that is asymptotically consistent. Our analysis can also be extended to a degree-corrected block model (DCSBM). In practice, the likelihood function can be estimated using more computationally efficient variational methods or consistent label estimation algorithms, allowing the criterion to be applied to large networks.

Full work available at URL: https://arxiv.org/abs/1502.02069

Recommendations

• On model selection for dense stochastic block models
• Model selection and clustering in stochastic block models based on the exact integrated complete data likelihood
• Hybrid maximum likelihood inference for stochastic block models
• Selective inference for latent block models
• Model selection in overlapping stochastic block models
• Empirical Bayes estimation for the stochastic blockmodel
• Consistency of maximum-likelihood and variational estimators in the stochastic block model
• Estimation and prediction for stochastic blockmodels for graphs with latent block structure
• Consistency of the maximum likelihood and variational estimators in a dynamic stochastic block model

zbMATH Keywords

likelihood ratio statisticmodel misspecificationnetwork communitiesstochastic block models

Mathematics Subject Classification ID

Asymptotic properties of parametric tests (62F05) Random graphs (graph-theoretic aspects) (05C80) Social networks; opinion dynamics (91D30)

Cited In (67)

• Two-sample test of stochastic block models
• A Time-Varying Network for Cryptocurrencies
• Heterogeneity pursuit for spatial point pattern with application to tree locations: a Bayesian semiparametric recourse
• Consistent model selection for the degree corrected stochastic blockmodel
• A spectral based goodness-of-fit test for stochastic block models
• A survey on theoretical advances of community detection in networks
• Network Estimation by Mixing: Adaptivity and More
• Community detection in complex networks: from statistical foundations to data science applications
• Recent advances on mechanisms of network generation: community, exchangeability, and scale-free properties
• Discussion of “Cocitation and Coauthorship Networks of Statisticians”
• Empirical Likelihood for Network Data
• On the Estimation of the Number of Communities for Sparse Networks
• PCABM: Pairwise Covariates-Adjusted Block Model for Community Detection
• Hypothesis testing for equality of latent positions in random graphs
• Modeling and Change Detection for Count-Weighted Multilayer Networks
• Two-sample test of stochastic block models via the maximum sampling entry-wise deviation
• Applications of dual regularized Laplacian matrix for community detection
• Universal rank inference via residual subsampling with application to large networks
• A survey on model-based co-clustering: high dimension and estimation challenges
• Large-scale estimation of random graph models with local dependence
• Inference for a generalised stochastic block model with unknown number of blocks and non-conjugate edge models
• A likelihood-ratio type test for stochastic block models with bounded degrees
• Hybrid maximum likelihood inference for stochastic block models
• Reliable prediction in the Markov stochastic block model
• Overlapping community detection in networks via sparse spectral decomposition
• Scalable estimation of epidemic thresholds via node sampling
• The Bethe Hessian and information theoretic approaches for online change-point detection in network data
• Optimal adaptivity of signed-polygon statistics for network testing
• Edgeworth expansions for network moments
• Mixture models and networks: The stochastic blockmodel
• Community detection by \(L_{0}\)-penalized graph Laplacian
• Asymptotically efficient estimators for stochastic blockmodels: the naive MLE, the rank-constrained MLE, and the spectral estimator
• Test on stochastic block model: local smoothing and extreme value theory
• Maximum likelihood estimation of sparse networks with missing observations
• Consistent Estimation of the Number of Communities via Regularized Network Embedding
• Spectral Clustering via Adaptive Layer Aggregation for Multi-Layer Networks
• Statistical embedding: beyond principal components
• Modeling the social media relationships of Irish politicians using a generalized latent space stochastic blockmodel
• Subsampling spectral clustering for stochastic block models in large-scale networks
• Model selection for Gaussian latent block clustering with the integrated classification likelihood
• Power enhancement and phase transitions for global testing of the mixed membership stochastic block model
• Network Structure Change Point Detection by Posterior Predictive Discrepancy
• On model selection for dense stochastic block models
• Title not available (Why is that?)
• Corrected Bayesian Information Criterion for Stochastic Block Models
• nett
• Adjusted chi-square test for degree-corrected block models
• Estimating the number of communities by spectral methods
• Network Cross-Validation for Determining the Number of Communities in Network Data
• Probabilistic Community Detection With Unknown Number of Communities
• Weighted stochastic block model
• Optimal Estimation of the Number of Network Communities
• Estimation and selection for the latent block model on categorical data
• Consistency and asymptotic normality of stochastic block models estimators from sampled data
• Hierarchical Community Detection by Recursive Partitioning
• Title not available (Why is that?)
• Extended stochastic block models with application to criminal networks
• Efficient split likelihood-based method for community detection of large-scale networks
• Vertex nomination: the canonical sampling and the extended spectral nomination schemes
• randnet
• Community detection with nodal information: likelihood and its variational approximation
• Consistent estimation of the number of communities in stochastic block models using cross-validation
• Fallacy of data-selective inference in modelling networks
• Statistical inference on group Rasch mixture network models
• Special invited paper: the SCORE normalization, especially for heterogeneous network and text data
• Using Maximum Entry-Wise Deviation to Test the Goodness of Fit for Stochastic Block Models
• Model-Based Clustering of Nonparametric Weighted Networks With Application to Water Pollution Analysis

This page was built for publication: Likelihood-based model selection for stochastic block models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q90065)