Exact Blind Community Detection From Signals on Multiple Graphs

被引:14
作者
Roddenberry, T. Mitchell [1 ]
Schaub, Michael T. [2 ,3 ]
Wai, Hoi-To [4 ]
Segarra, Santiago [1 ]
机构
[1] Rice Univ, Dept Elect & Comp Engn, Houston, TX 77005 USA
[2] Univ Oxford, Dept ES, Oxford OX1 2JD, England
[3] Rhein Westfal TH Aachen, Dept CS, D-52062 Aachen, Germany
[4] Chinese Univ Hong Kong, Dept Syst Engn & Engn Management, Shatin, Hong Kong, Peoples R China
基金
欧盟地平线“2020”;
关键词
Symmetric matrices; Signal processing algorithms; Image edge detection; Laplace equations; Clustering algorithms; Signal processing; Mathematical model; Blind system identification; community detection; graph signal processing; network process; spectral clustering; IDENTIFICATION; INFERENCE; CONSENSUS; MODELS;
D O I
10.1109/TSP.2020.3016494
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
Networks and data supported on graphs have become ubiquitous in the sciences and engineering. This paper studies the 'blind' community detection problem, where we seek to infer the community structure of a graph model given the observation of independent graph signals on a set of nodes whose connections are unknown. We model each observation as filtered white noise, where the underlying network structure varies with every observation. These varying network structures are modeled as independent realizations of a latent planted partition model (PPM), justifying our assumption of a constant underlying community structure over all observations. Under certain conditions on the graph filter and PPM parameters, we suggest simple algorithms for determining (i) the number of latent communities and (ii) the associated partitions of the PPM. We then prove statistical guarantees in the asymptotic and non-asymptotic sampling cases. Numerical experiments on real and synthetic data demonstrate the efficacy of these algorithms.
引用
收藏
页码:5016 / 5030
页数:15
相关论文
共 56 条
[1]   Community detection and stochastic block models: Recent developments [J].
Abbe, Emmanuel .
Journal of Machine Learning Research, 2018, 18 :1-86
[2]  
[Anonymous], 2012, Matrix Analysis
[3]   Proximal-Gradient Algorithms for Tracking Cascades Over Social Networks [J].
Baingana, Brian ;
Mateos, Gonzalo ;
Giannakis, Georgios B. .
IEEE JOURNAL OF SELECTED TOPICS IN SIGNAL PROCESSING, 2014, 8 (04) :563-575
[4]   Efficient and principled method for detecting communities in networks [J].
Ball, Brian ;
Karrer, Brian ;
Newman, M. E. J. .
PHYSICAL REVIEW E, 2011, 84 (03)
[5]   COVARIANCE REGULARIZATION BY THRESHOLDING [J].
Bickel, Peter J. ;
Levina, Elizaveta .
ANNALS OF STATISTICS, 2008, 36 (06) :2577-2604
[6]  
Boutsidis C, 2015, PR MACH LEARN RES, V37, P40
[7]   On the sample covariance matrix estimator of reduced effective rank population matrices, with applications to fPCA [J].
Bunea, Florentina ;
Xiao, Luo .
BERNOULLI, 2015, 21 (02) :1200-1230
[8]   Inference of Gene Regulatory Networks with Sparse Structural Equation Models Exploiting Genetic Perturbations [J].
Cai, Xiaodong ;
Bazerque, Juan Andres ;
Giannakis, Georgios B. .
PLOS COMPUTATIONAL BIOLOGY, 2013, 9 (05)
[9]  
Cormen T. H., 2009, Introduction To Algorithms
[10]   Consistent resting-state networks across healthy subjects [J].
Damoiseaux, J. S. ;
Rombouts, S. A. R. B. ;
Barkhof, F. ;
Scheltens, P. ;
Stam, C. J. ;
Smith, S. M. ;
Beckmann, C. F. .
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA, 2006, 103 (37) :13848-13853