A Useful Criterion on Studying Consistent Estimation in Community Detection
Abstract
:1. Introduction
1.1. Spectral Clustering Approaches
1.2. Separation Condition, Alternative Separation Condition and Sharp Threshold
1.3. Inconsistencies on Separation Condition in Some Previous Works
1.4. Our Findings
- (i)
- We summarize the idea of using the separation condition of a standard network and sharp threshold of the ER random graph to study the consistent estimations of different spectral methods designed via eigen-decomposition or singular value decomposition of the adjacency matrix or its variants under different models that can degenerate to SBM under mild conditions as a four-step criterion, SCSTC. The separation condition is used to study the consistency of the theoretical upper bound for the spectral method, and the sharp threshold can be used to study the network sparsity. The theoretical results of upper bounds for different spectral methods can be compared by SCSTC. Using this criterion, a few inconsistent phenomenons of some previous works are found.
- (ii)
- Under MMSB and DCMM, we study the consistencies of the SPACL algorithm proposed in [44] and its extended version using the recent techniques on row-wise eigenvector deviation developed in [64,65]. Compared with the original results of [43,44], our main theoretical results enjoy smaller error rates by lesser dependence on K and . Meanwhile, our main theoretical results have weaker requirements on the network sparsity and the lower bound of the smallest nonzero singular value of the population adjacency matrix. For details, see Table 3 and Table 4.
- (iii)
- Our results for DCMM are consistent with those for MMSB when DCMM degenerates to MMSB under mild conditions. Using SCSTC, under mild conditions, our main theoretical results under DCMM are consistent with those of [41]. This answers the question that the phenomenon that the main results of [43,44] do not match those of [41] occurs due to the fact that in Refs. [43,44], the theoretical results of error rates are sub-optimal. We also find that our theoretical results (as well as those of [41]) under both MMSB and DCMM match the classical results on the separation condition and sharp threshold, i.e., achieve thresholds in Equations (1)–(3). Using the bound of instead of to establish the upper bound of error rate under SBM in [30], the two spectral methods studied in [30] achieve thresholds in Equations (1)–(3), which answers the question of why the separation condition obtained from error rate of [41] does not match that obtained from the error rate of [30]. Using or influences the row-wise eigenvector deviations in Theorem 3.1 of [44] and Theorem I.3 of [43], and thus using or influences the separation conditions and sharp thresholds of [43,44]. For comparison, our bound on row-wise eigenvector deviation is obtained by using the techniques developed in [64,65] and that of [41] is obtained by applying the modified Theorem 2.1 of [66]; therefore, using or has no influence on the separation conditions and sharp thresholds of ours and that of [41]. For details, see Table 1 and Table 2. In a word, using SCSTC, the spectral methods proposed and studied in [26,30,41,43,44,48,49,50,67,68] or some other spectral methods fitting models that can reduce to achieve thresholds in Equations (1)–(3).
- (iv)
- We verify our threshold in Equation (2) by some computer-generated networks in Section 6. The numerical results for networks generated under when and show that SPACL and its extended version achieve a threshold in Equation (2), and results for networks generated from when and show that the spectral methods considered in [26,30,48,50] achieve the threshold in Equation (2).
2. Mixed Membership Stochastic Blockmodel
- (I1) .
- (I2) There is at least one pure node for each of the K communities.
- Let be the top-K eigen-decomposition of such that .
- Run SP algorithm on the rows of U assuming that there are K communities to obtain .
- Set .
- Recover by setting for .
Algorithm 1 SPACL [44] |
|
3. Consistency under MMSB
- (A1)
- .
- We emphasize that the bound of Theorem 3.1 of [44] should be instead of for where the function is defined in Equation (7) of [44], and this is also pointed out by Table 2 of [63]. The reason is that in the proof part of Theorem 3.1 [44], from step (iii) to step (iv), they should keep the term since this term is much larger than 1. We can also find that the bound in Theorem 3.1 [44] should multiply from Theorem VI.1 [44] directly. For comparison, this bound is times our bound in Lemma 2. Meanwhile, by the proof of the bound in Theorem 3.1 of [44], we see that the bound depends on the upper bound of , and [44] applies Theorem 5.2 of [30] such that with high probability. Since is the upper bound of the difference between a regularization of A and . Therefore, if we are only interested in bounding instead of , the upper bound of Theorem 3.1 [44] should be , which is at least times our bound in Lemma 2. Furthermore, the upper bound of the row-wise eigenspace error in Lemma 2 does not rely on the upper bound of as long as holds. Therefore, whether using or does not change the bound in Lemma 2.
- Since by basic algebra, the lower bound requirement on in Assumption 3.1 of [44] gives that , which suggests that Theorem 3.1 [44] requires , and this also matches with the requirement on in Theorem VI.1 of [44] (and this is also pointed out by Table 1 of [63]). For comparison, our requirement on sparsity given in Assumption (A1) is , which is weaker than . Similarly, in our Lemma 2, the requirement gives , thus we have which is consistent with Assumption (A1).
4. Separation Condition and Sharp Threshold Criterion
- (a) By Corollary 1, we know that should shrink slower than for consistent estimation. Therefore, the separation condition should shrink slower than (i.e., Equation (1)), and this threshold is consistent with Corollary 1 of [59] and Equation (17) of [49]. The alternative separation condition should shrink slower than 1 (i.e., Equation (2)).
- (c) By Remark 5, using , we know that in Ref. [44], Equation (3) is , so should shrink slower than . Thus, for [44], the separation condition is , and the alternative separation condition is , which are sub-optimal compared with ours in (a). Using , and Equation (3) in Ref. [44], which is , we see that for [44], now the separation condition is and the alternative separation condition is .
- (d) For comparison, the error bound of Corollary 3.2 [30] built under SBM for community detection is for with , so should shrink slower than . Thus the separation condition for [30] is . However, as we analyzed in the first bullet given after Lemma 2, [30] applied to build their consistency results. Instead, we apply to the built theoretical results of [30], and the error bound of Corollary 3.2 [30] is , which returns the same separation condition as our Corollary 1 and Theorem 2.2 of [41] now. Following a similar analysis to (a)–(c), we can obtain an alternative separation condition for [30] immediately, and the results are provided in Table 2. Meanwhile, as analyzed in the first bullet given after Lemma 2, whether using or does not change our error rates. By carefully analyzing the proof of 2.1 of [41], we see that whether using or also does not change their row-wise large deviation, hence it does not influence their upper bound of the error rate for their Mixed-SCORE.
- Check whether the theoretical upper bound of the error rate contains (note that is probability matrix and maximum entries of should be set as 1), where the separation parameter always appears when considering the lower bound of . If it contains , move to the next step. Otherwise, it suggests possible improvements for the consistency by considering in the proofs.
- Let and network degenerate to the standard network whose numbers of nodes in each community are in the same order and can been seen as (i.e., a with in the case of a non-overlapping network or a with in the case of an overlapping network, and we will mainly focus on with for convenience.). Let the model degenerate to with , and then we obtain the new theoretical upper bound of the error rate. Note that if the model does consider degree heterogeneity, sparsity parameter should be considered in the theoretical upper bound of error rate in . If the model considers degree heterogeneity, when it degenerates to with , appears at this step. Meanwhile, if is not contained in the error rate of when the model does not consider degree heterogeneity, it suggests possible improvements by considering .
- Let be the probability matrix when the model degenerates to such that P has diagonal entries and non-diagonal entries . So, and separation condition since the maximum entry of is assumed to be 1. Compute the lower bound requirement of for consistency estimation through analyzing the new bound obtained in . Compute separation condition using the lower bound requirement for . The sharp threshold for the ER random graph is obtained from the lower bound requirement on for the consistency estimation under the setting that and .
- Compare the separation condition and the sharp threshold obtained in with Equations (1) and (3), respectively. If the sharp threshold or the separation condition , then this leaves improvements on the requirement of the network sparsity or theoretical upper bound of the error rate. If the sharp threshold is and the separation condition is , the optimality of the theoretical results on both error rates and the requirement of network sparsity is guaranteed. Finally, if the sharp threshold or separation condition , this suggests that the theoretical result is obtained based on instead of .
- In , we give a few examples. When applying SCSTC to the main results of [40,48,67], we stop at as analyzed in Remark 8, suggesting possible improvements by considering for these works. Meanwhile, for the theoretical result without considering , we can also move to to obtain the new theoretical upper bound of the error rate, which is related with ρ and n. Discussions on the theoretical upper bounds of error rates of [50,68] given in Remark 8 are examples of this case.
- In , letting and the model reduce to for the non-overlapping network or for the overlapping network can always simplify the theoretical upper bound of error rate, as shown by our Corollaries 1 and 2. Here, we provide some examples about how to make a model degenerate to SBM. For in this paper, when all nodes are pure, MMSB degenerates to SBM; for the model introduced in Section 5 or DCSBM considered in [30,48,50], setting makes DCMM and DCSBM degenerates to SBM when all nodes are pure, similar to the ScBM and DCScBM considered in [67,68,71], the OCCAM model of [40], the stochastic blockmodel with the overlap proposed in [46], the extensions of SBM and DCSBM for hypergraph networks considered in [73,74,75], and so forth.
- In and , the separation condition can be replaced by an alternative separation condition.
- When using SCSTC to build and compare theoretical results for the spectral clustering method, the key point is computing the lower bound for when the probability matrix P has diagonal entries and non-diagonal entries from the theoretical upper bound of the error rate for a certain spectral method. If this lower bound is consistent with that of Equation (1), this suggests theoretical optimality, and otherwise it suggests possible improvements by following the four steps of SCSTC.
- Theorem 4.4 of [48] proposes the upper bound of the error rate for their regularized spectral clustering (RSC) algorithm, designed based on a regularized Laplacian matrix under DCSBM. However, since [48] does not study the lower bound (in the [48] language) of and m, we cannot directly obtain the separation condition from their main theorem. Meanwhile, the main result of [48] does not consider the requirement on the network sparsity, which leaves some improvements. Ref. [48] does not study the theoretical optimal choice for the RSC regularizer τ. After considering and sparsity parameter ρ, one can obtain the theoretical optimal choice for τ, and this is helpful for explaining and choosing the empirical optimal choice for τ. Therefore, the feasible network implementation of SCSTC is obtaining the theoretical optimal choices for some tuning parameters, such as regularizer τ of the RSC algorithm. By using SCSTC, we can find that RSC achieves thresholds in Equations (1)–(3), and we omit proofs for it in this paper.
- Refs. [26,49] study two algorithms designed based on the Laplacian matrix and its regularized version under SBM. They obtain meaningful results, but do not consider the network sparsity parameter ρ and separation parameter . After obtaining improved error bounds which are consistent with separation condition using SCSTC, one can also obtain the theoretical optimal choice for regularizer τ of the RSC-τ algorithm considered in [49] and find that the two algorithms considered in [26,49] achieve thresholds in Equations (1)–(3).
- Theorem 2.2 of [50] provides an upper bound of their SCORE algorithm under DCSBM. However, since they do not consider the influence of , we cannot directly obtain the separation condition from their main result. Meanwhile, by setting their , DCSBM degenerates to SBM, which gives that their by their assumption Equation (2.9). Hence, when , the upper bound of Theorem 2.2 in [50] is . The upper bound of error rate in Corollary 3.2 of [30] is when using under the setting that and . We see that grows faster than , which suggests that there is space to improve the main result of [50] in the aspects of the separation condition and error rates. Furthermore, using SCSTC, we can find that SCORE achieves thresholds in Equations (1)–(3) because its extension mixed-SCORE [41] achieves thresholds in Equations (1)–(3).
- Ref. [67] proposes two models, ScBM and DCScBM, to model the directed networks and an algorithm DI-SIM based on the directed regularized Laplacian matrix to fit DCScBM. However, similar to [48], their main theoretical result in their Theorem C.1 does not consider the lower bound of (in the language of Ref. [67]) and , which causes that we cannot obtain the separation condition when DCScBM degenerates to SBM. Meanwhile, their Theorem C.1 also lacks a lower bound requirement on network sparsity. Hence, there is space to improve the theoretical guarantees of [67]. Similar to [48,49], we can also obtain the theoretical optimal choices for regularizer τ of the DI-SIM algorithm and prove that DI-SIM achieves the thresholds in Equations (1)–(3) since it is the directed version of RSC [48].
- Ref. [68] mainly studies the theoretical guarantee for the D-SCORE algorithm proposed by [14] to fit a special case of the DCScBM model for directed networks. By setting their for , their directed-DCBM degenerates to SBM. Meanwhile, since their , their mis-clustering rate is , which matches that of [30] under SBM when setting as a constant. However, if setting as , then the error rate is , which is sub-optimal compared with that of [30]. Meanwhile, similar to [50,68], the main result does not consider the influences of K and , causing a lack of a separation condition. Hence, the main results of [68] can be improved by considering K, , or a more optimal choice of to make their main results comparable with those of [30] when directed-DCBM degenerates to SBM. Using SCSTC, we can find that the D-SCORE also achieves thresholds in Equations (1)–(3) since it is the directed version of SCORE [50].
5. Degree Corrected Mixed Membership Model
- (II1) and has unit diagonals.
- (II2) There is at least one pure node for each of the K communities.
- Let be the top-K eigen-decomposition of such that . Let , where is a diagonal matrix whose i-th diagonal entry is for .
- Run SVM-cone algorithm on assuming that there are K communities to obtain .
- Set .
- Recover by setting for .
Algorithm 2 SVM-cone-DCMMSB [43] |
|
Consistency under DCMM
- (A2)
- .
6. Numerical Results
- (a)
- Set .
- (b)
- Let W be an symmetric matrix such that all diagonal entries of W are 0, and are independent centered Bernoulli with parameters . Let be the adjacency matrix of a simulated network with mixed memberships under MMSB (so there are no loops).
- (c)
- Apply spectral clustering method to A with K communities. Record the error rate.
- (d)
- Repeat (b)–(c) 50 times, and report the mean of the error rates over the 50 times.
- Obtain the graph Laplacian , where D is a diagonal matrix with for .
- Obtain , the top K eigen-decomposition of L.
- Apply k-means algorithm to to obtain .
- Obtain , the top K eigen-decomposition of A.
- Apply k-means algorithm to to obtain .
- Obtain the regularized graph Laplacian , where , and the default is the average node degree.
- Obtain , the top K eigen-decomposition of . Let be the row-normalized version of .
- Apply k-means algorithm to to obtain .
- Obtain the K (unit-norm) leading eigenvectors of A: .
- Obtain an matrix such that for ,
- Apply k-means algorithm to to obtain .
7. Conclusions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
Abbreviations
SCSTC | separation condition and sharp threshold criterion |
SBM | stochastic blockmodels |
DCSBM | degree corrected stochastic blockmodel |
MMSB | mixed membership stochastic blockmodel |
DCMM | degree corrected mixed membership model |
SBMO | stochastic blockmodel with overlap |
OCCAM | overlapping continuous community assignment model |
RSC | regularized spectral clustering |
SCORE | spectral clustering on ratios-of-eigenvectors |
SPACL | sequential projection after cleaning |
ER | Erdös–Rényi |
IS | ideal simplex |
IC | ideal cone |
SP | successive projection algorithm |
oPCA | ordinary principle component analysis |
nPCA | normalized principle component analysis |
Appendix A. Additional Experiments
Appendix B. Vertex Hunting Algorithms
Algorithm A1 Successive projection (SP) [51] |
|
Algorithm A2 SVM-cone [43] |
|
Appendix C. Proof of Consistency under MMSB
Appendix C.1. Proof of Lemma 1
Appendix C.2. Proof of Lemma 2
Appendix C.3. Proof of Theorem 1
Appendix C.4. Proof of Corollary 1
Appendix D. Proof of Consistency under DCMM
Appendix D.1. Proof of Lemma 3
Appendix D.2. Proof of Lemma 4
Appendix D.3. Proof of Lemma 5
Appendix D.4. Proof of Theorem 2
Appendix D.5. Proof of Corollary 2
Appendix D.6. Basic Properties of Ω under DCMM
Appendix D.7. Bounds between Ideal SVM-cone-DCMMSB and SVM-cone-DCMMSB
- We bound first. Set for convenience. For , we haveNow we aim to bound . For convenience, set . We haveThen, we have
- for , since , we have
- for , recall that , we have
- for , we provide some simple facts first: . Since is the best rank K approximation to A in the spectral norm, and therefore since with rank K and can also be viewed as a rank K approximation to A. This leads to . By Lemma H.2 [43], . by the lower bound requirement of in Lemma 5, and we also have . For , let for convenience. Based on the above facts and Lemma A2, we haveRecall that , we have where the last inequality holds by Lemma A1. Similarly, we have where the last inequality holds by the proof of Lemma 5. Then we haveCombining the above results, we have
References
- Watts, D.J.; Strogatz, S.H. Collective dynamics of ‘small-world’networks. Nature 1998, 393, 440–442. [Google Scholar] [CrossRef] [PubMed]
- Newman, M.E. Scientific collaboration networks. II. Shortest paths, weighted networks, and centrality. Phys. Rev. E 2001, 64, 016132. [Google Scholar] [CrossRef] [PubMed]
- Dunne, J.A.; Williams, R.J.; Martinez, N.D. Food-web structure and network theory: The role of connectance and size. Proc. Natl. Acad. Sci. USA 2002, 99, 12917–12922. [Google Scholar] [CrossRef]
- Newman, M.E.J. Coauthorship networks and patterns of scientific collaboration. Proc. Natl. Acad. Sci. USA 2004, 101, 5200–5205. [Google Scholar] [CrossRef]
- Notebaart, R.A.; van Enckevort, F.H.; Francke, C.; Siezen, R.J.; Teusink, B. Accelerating the reconstruction of genome-scale metabolic networks. BMC Bioinform. 2006, 7, 296. [Google Scholar] [CrossRef]
- Pizzuti, C. Ga-net: A genetic algorithm for community detection in social networks. In International Conference on Parallel Problem Solving from Nature; Springer: Berlin/Heidelberg, Germany, 2008; pp. 1081–1090. [Google Scholar]
- Jackson, M.O. Social and Economic Networks; Princeton University Press: Princeton, NJ, USA, 2010. [Google Scholar]
- Gao, J.; Liang, F.; Fan, W.; Wang, C.; Sun, Y.; Han, J. On community outliers and their efficient detection in information networks. In Proceedings of the 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Washington, DC, USA, 24–28 July 2010; pp. 813–822. [Google Scholar]
- Rubinov, M.; Sporns, O. Complex network measures of brain connectivity: Uses and interpretations. Neuroimage 2010, 52, 1059–1069. [Google Scholar] [CrossRef]
- Su, G.; Kuchinsky, A.; Morris, J.H.; States, D.J.; Meng, F. GLay: Community structure analysis of biological networks. Bioinformatics 2010, 26, 3135–3137. [Google Scholar] [CrossRef]
- Lin, W.; Kong, X.; Yu, P.S.; Wu, Q.; Jia, Y.; Li, C. Community detection in incomplete information networks. In Proceedings of the 21st International Conference on World Wide Web, Lyon, France, 16–20 April 2012; pp. 341–350. [Google Scholar]
- Scott, J.; Carrington, P.J. The SAGE Handbook of Social Network Analysis; SAGE Publications: London, UK, 2014. [Google Scholar]
- Bedi, P.; Sharma, C. Community detection in social networks. Wiley Interdiscip. Rev. Data Min. Knowl. Discov. 2016, 6, 115–135. [Google Scholar] [CrossRef]
- Ji, P.; Jin, J. Coauthorship and citation networks for statisticians. Ann. Appl. Stat. 2016, 10, 1779–1812. [Google Scholar] [CrossRef]
- Ji, P.; Jin, J.; Ke, Z.T.; Li, W. Co-citation and Co-authorship Networks of Statisticians. J. Bus. Econ. Stat. 2022, 40, 469–485. [Google Scholar] [CrossRef]
- Newman, M.E. The structure and function of complex networks. SIAM Rev. 2003, 45, 167–256. [Google Scholar] [CrossRef]
- Newman, M.E.; Girvan, M. Finding and evaluating community structure in networks. Phys. Rev. E 2004, 69, 026113. [Google Scholar] [CrossRef] [PubMed]
- Boccaletti, S.; Latora, V.; Moreno, Y.; Chavez, M.; Hwang, D.U. Complex networks: Structure and dynamics. Phys. Rep. 2006, 424, 175–308. [Google Scholar] [CrossRef]
- Fortunato, S. Community detection in graphs. Phys. Rep. 2010, 486, 75–174. [Google Scholar] [CrossRef]
- Fortunato, S.; Hric, D. Community detection in networks: A user guide. Phys. Rep. 2016, 659, 1–44. [Google Scholar] [CrossRef]
- Abbe, E.; Bandeira, A.S.; Hall, G. Exact Recovery in the Stochastic Block Model. IEEE Trans. Inf. Theory 2016, 62, 471–487. [Google Scholar] [CrossRef]
- Fortunato, S.; Newman, M.E. 20 years of network community detection. Nat. Phys. 2022, 1–3. [Google Scholar] [CrossRef]
- Goldenberg, A.; Zheng, A.X.; Fienberg, S.E.; Airoldi, E.M. A survey of statistical network models. Found. Trends Mach. Learn. 2010, 2, 129–233. [Google Scholar] [CrossRef]
- Holland, P.W.; Laskey, K.B.; Leinhardt, S. Stochastic blockmodels: First steps. Soc. Netw. 1983, 5, 109–137. [Google Scholar] [CrossRef]
- Snijders, T.A.; Nowicki, K. Estimation and prediction for stochastic blockmodels for graphs with latent block structure. J. Classif. 1997, 14, 75–100. [Google Scholar] [CrossRef]
- Rohe, K.; Chatterjee, S.; Yu, B. Spectral clustering and the high-dimensional stochastic blockmodel. Ann. Stat. 2011, 39, 1878–1915. [Google Scholar] [CrossRef]
- Choi, D.S.; Wolfe, P.J.; Airoldi, E.M. Stochastic blockmodels with a growing number of classes. Biometrika 2012, 99, 273–284. [Google Scholar] [CrossRef]
- Sussman, D.L.; Tang, M.; Fishkind, D.E.; Priebe, C.E. A consistent adjacency spectral embedding for stochastic blockmodel graphs. J. Am. Stat. Assoc. 2012, 107, 1119–1128. [Google Scholar] [CrossRef]
- Latouche, P.; Birmelé, E.; Ambroise, C. Model selection in overlapping stochastic block models. Electron. J. Stat. 2014, 8, 762–794. [Google Scholar] [CrossRef]
- Lei, J.; Rinaldo, A. Consistency of spectral clustering in stochastic block models. Ann. Stat. 2015, 43, 215–237. [Google Scholar] [CrossRef]
- Sarkar, P.; Bickel, P.J. Role of normalization in spectral clustering for stochastic blockmodels. Ann. Stat. 2015, 43, 962–990. [Google Scholar] [CrossRef]
- Lyzinski, V.; Tang, M.; Athreya, A.; Park, Y.; Priebe, C.E. Community detection and classification in hierarchical stochastic blockmodels. IEEE Trans. Netw. Sci. Eng. 2016, 4, 13–26. [Google Scholar] [CrossRef]
- Valles-Catala, T.; Massucci, F.A.; Guimera, R.; Sales-Pardo, M. Multilayer stochastic block models reveal the multilayer structure of complex networks. Phys. Rev. X 2016, 6, 011036. [Google Scholar] [CrossRef]
- Lei, J. A goodness-of-fit test for stochastic block models. Ann. Stat. 2016, 44, 401–424. [Google Scholar] [CrossRef]
- Tabouy, T.; Barbillon, P.; Chiquet, J. Variational inference for stochastic block models from sampled data. J. Am. Stat. Assoc. 2020, 115, 455–466. [Google Scholar] [CrossRef]
- Airoldi, E.M.; Blei, D.M.; Fienberg, S.E.; Xing, E.P. Mixed Membership Stochastic Blockmodels. J. Mach. Learn. Res. 2008, 9, 1981–2014. [Google Scholar] [PubMed]
- Wang, F.; Li, T.; Wang, X.; Zhu, S.; Ding, C. Community discovery using nonnegative matrix factorization. Data Min. Knowl. Discov. 2011, 22, 493–521. [Google Scholar] [CrossRef]
- Airoldi, E.M.; Wang, X.; Lin, X. Multi-way blockmodels for analyzing coordinated high-dimensional responses. Ann. Appl. Stat. 2013, 7, 2431–2457. [Google Scholar] [CrossRef] [PubMed]
- Panov, M.; Slavnov, K.; Ushakov, R. Consistent Estimation of Mixed Memberships with Successive Projections. In International Conference on Complex Networks and Their Applications; Springer: Cham, Switzerland, 2017; pp. 53–64. [Google Scholar]
- Zhang, Y.; Levina, E.; Zhu, J. Detecting overlapping communities in networks using spectral methods. SIAM J. Math. Data Sci. 2020, 2, 265–283. [Google Scholar] [CrossRef]
- Jin, J.; Ke, Z.T.; Luo, S. Estimating network memberships by simplex vertex hunting. arXiv 2017, arXiv:1708.07852. [Google Scholar]
- Mao, X.; Sarkar, P.; Chakrabarti, D. On Mixed Memberships and Symmetric Nonnegative Matrix Factorizations. In Proceedings of the 34th International Conference of Machine Learning, Sydney, Australia, 6–11 August 2017; pp. 2324–2333. [Google Scholar]
- Mao, X.; Sarkar, P.; Chakrabarti, D. Overlapping Clustering Models, and One (class) SVM to Bind Them All. In Proceedings of the Advances in Neural Information Processing Systems, Montréal, QC, Canada, 3–8 December 2018; Volume 31, pp. 2126–2136. [Google Scholar]
- Mao, X.; Sarkar, P.; Chakrabarti, D. Estimating Mixed Memberships With Sharp Eigenvector Deviations. J. Am. Stat. Assoc. 2020, 116, 1928–1940. [Google Scholar] [CrossRef]
- Karrer, B.; Newman, M.E.J. Stochastic blockmodels and community structure in networks. Phys. Rev. E 2011, 83, 16107. [Google Scholar] [CrossRef]
- Kaufmann, E.; Bonald, T.; Lelarge, M. A spectral algorithm with additive clustering for the recovery of overlapping communities in networks. Theor. Comput. Sci. 2017, 742, 3–26. [Google Scholar] [CrossRef]
- Von Luxburg, U. A tutorial on spectral clustering. Stat. Comput. 2007, 17, 395–416. [Google Scholar] [CrossRef]
- Qin, T.; Rohe, K. Regularized spectral clustering under the degree-corrected stochastic blockmodel. Adv. Neural Inf. Process. Syst. 2013, 26, 3120–3128. [Google Scholar]
- Joseph, A.; Yu, B. Impact of regularization on spectral clustering. Ann. Stat. 2016, 44, 1765–1791. [Google Scholar] [CrossRef]
- Jin, J. Fast community detection by SCORE. Ann. Stat. 2015, 43, 57–89. [Google Scholar] [CrossRef]
- Gillis, N.; Vavasis, S.A. Semidefinite Programming Based Preconditioning for More Robust Near-Separable Nonnegative Matrix Factorization. SIAM J. Optim. 2015, 25, 677–698. [Google Scholar] [CrossRef]
- Mossel, E.; Neeman, J.; Sly, A. Consistency thresholds for binary symmetric block models. arXiv 2014, arXiv:1407.1591. [Google Scholar]
- Abbe, E. Community detection and stochastic block models: Recent developments. J. Mach. Learn. Res. 2017, 18, 6446–6531. [Google Scholar]
- Hajek, B.; Wu, Y.; Xu, J. Achieving Exact Cluster Recovery Threshold via Semidefinite Programming: Extensions. IEEE Trans. Inf. Theory 2016, 62, 5918–5937. [Google Scholar] [CrossRef]
- Agarwal, N.; Bandeira, A.S.; Koiliaris, K.; Kolla, A. Multisection in the Stochastic Block Model using Semidefinite Programming. arXiv 2017, arXiv:1507.02323. [Google Scholar]
- Bandeira, A.S. Random Laplacian Matrices and Convex Relaxations. Found. Comput. Math. 2018, 18, 345–379. [Google Scholar] [CrossRef]
- Abbe, E.; Sandon, C. Community Detection in General Stochastic Block models: Fundamental Limits and Efficient Algorithms for Recovery. In Proceedings of the 2015 IEEE 56th Annual Symposium on Foundations of Computer Science, Berkeley, CA, USA, 17–20 October 2015; pp. 670–688. [Google Scholar]
- Gao, C.; Ma, Z.; Zhang, A.Y.; Zhou, H.H. Achieving Optimal Misclassification Proportion in Stochastic Block Models. J. Mach. Learn. Res. 2017, 18, 1–45. [Google Scholar]
- McSherry, F. Spectral partitioning of random graphs. In Proceedings of the 2001 IEEE International Conference on Cluster Computing, Newport Beach, CA, USA, 8–11 October 2001; pp. 529–537. [Google Scholar]
- Newman, M.E. Assortative mixing in networks. Phys. Rev. Lett. 2002, 89, 208701. [Google Scholar] [CrossRef]
- Erdös, P.; Rényi, A. On the evolution of random graphs. In The Structure and Dynamics of Networks; Princeton University Press: Princeton, NJ, USA, 2011; pp. 38–82. [Google Scholar] [CrossRef]
- Blum, A.; Hopcroft, J.; Kannan, R. Foundations of Data Science; Number 1; Cambridge University Press: Cambridge, UK, 2020; pp. 1–465. [Google Scholar]
- Lei, L. Unified ℓ2→∞ Eigenspace Perturbation Theory for Symmetric Random Matrices. arXiv 2019, arXiv:1909.04798. [Google Scholar]
- Chen, Y.; Chi, Y.; Fan, J.; Ma, C. Spectral methods for data science: A statistical perspective. Found. Trends Mach. Learn. 2021, 14, 566–806. [Google Scholar] [CrossRef]
- Cape, J.; Tang, M.; Priebe, C.E. The two-to-infinity norm and singular subspace geometry with applications to high-dimensional statistics. Ann. Stat. 2019, 47, 2405–2439. [Google Scholar] [CrossRef]
- Abbe, E.; Fan, J.; Wang, K.; Zhong, Y. Entrywise Eigenvector Analysis of Random Matrices with Low Expected Rank. Ann. Stat. 2020, 48, 1452–1474. [Google Scholar] [CrossRef] [PubMed]
- Rohe, K.; Qin, T.; Yu, B. Co-clustering directed graphs to discover asymmetries and directional communities. Proc. Natl. Acad. Sci. USA 2016, 113, 12679–12684. [Google Scholar] [CrossRef] [PubMed]
- Wang, Z.; Liang, Y.; Ji, P. Spectral Algorithms for Community Detection in Directed Networks. J. Mach. Learn. Res. 2020, 21, 1–45. [Google Scholar]
- Cai, T.T.; Li, X. Robust and computationally feasible community detection in the presence of arbitrary outlier nodes. Ann. Stat. 2015, 43, 1027–1059. [Google Scholar] [CrossRef]
- Tropp, J.A. User-Friendly Tail Bounds for Sums of Random Matrices. Found. Comput. Math. 2012, 12, 389–434. [Google Scholar] [CrossRef]
- Zhou, Z.; A.Amini, A. Analysis of spectral clustering algorithms for community detection: The general bipartite setting. J. Mach. Learn. Res. 2019, 20, 1–47. [Google Scholar]
- Zhao, Y.; Levina, E.; Zhu, J. Consistency of community detection in networks under degree-corrected stochastic block models. Ann. Stat. 2012, 40, 2266–2292. [Google Scholar] [CrossRef]
- Ghoshdastidar, D.; Dukkipati, A. Consistency of Spectral Partitioning of Uniform Hypergraphs under Planted Partition Model. In Proceedings of the Advances in Neural Information Processing Systems 27, Montreal, QC, Canada, 8–13 December 2014; Volume 27, pp. 397–405. [Google Scholar]
- Ke, Z.T.; Shi, F.; Xia, D. Community Detection for Hypergraph Networks via Regularized Tensor Power Iteration. arXiv 2019, arXiv:1909.06503. [Google Scholar]
- Cole, S.; Zhu, Y. Exact recovery in the hypergraph stochastic block model: A spectral algorithm. Linear Algebra Its Appl. 2020, 593, 45–73. [Google Scholar] [CrossRef]
- Bandeira, A.S.; van Handel, R. Sharp nonasymptotic bounds on the norm of random matrices with independent entries. Ann. Probab. 2016, 44, 2479–2506. [Google Scholar] [CrossRef]
- Cape, J. Orthogonal Procrustes and norm-dependent optimality. Electron. J. Linear Algebra 2020, 36, 158–168. [Google Scholar] [CrossRef]
Model | Separation Condition | Sharp Threshold | |
---|---|---|---|
Ours using | MMSB&DCMM | ||
Ours using | MMSB&DCMM | ||
Ref. [41] using (original) | DCMM | ||
Ref. [41] using | DCMM | ||
Refs. [43,44] using (original) | MMSB&DCMM | ||
Refs. [43,44] using | MMSB&DCMM | ||
Ref. [30] using (original) | SBM&DCSBM | ||
Ref. [30] using | SBM&DCSBM |
Model | Alternative Separation Condition | |
---|---|---|
Ours using | MMSB&DCMM | 1 |
Ours using | MMSB&DCMM | 1 |
Ref. [41] using (original) | DCMM | 1 |
Ref. [41] using | DCMM | 1 |
Refs. [43,44] using (original) | MMSB&DCMM | |
Refs. [43,44] using | MMSB&DCMM | |
Ref. [30] using (original) | SBM&DCSBM | |
Ref. [30] using | SBM&DCSBM | 1 |
Dependence on K | Dependence on | ||||
---|---|---|---|---|---|
Ours | |||||
[44] |
Dependence on K | Dependence on | |||||
---|---|---|---|---|---|---|
Ours | arbitrary | |||||
[43] | from Dirichlet |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2022 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Qing, H. A Useful Criterion on Studying Consistent Estimation in Community Detection. Entropy 2022, 24, 1098. https://doi.org/10.3390/e24081098
Qing H. A Useful Criterion on Studying Consistent Estimation in Community Detection. Entropy. 2022; 24(8):1098. https://doi.org/10.3390/e24081098
Chicago/Turabian StyleQing, Huan. 2022. "A Useful Criterion on Studying Consistent Estimation in Community Detection" Entropy 24, no. 8: 1098. https://doi.org/10.3390/e24081098
APA StyleQing, H. (2022). A Useful Criterion on Studying Consistent Estimation in Community Detection. Entropy, 24(8), 1098. https://doi.org/10.3390/e24081098