A Nearly-Linear Time Algorithm for Exact Community Recovery in Stochastic Block Model

Peng Wang, Zirui Zhou, Anthony Man-Cho So
Proceedings of the 37th International Conference on Machine Learning, PMLR 119:10126-10135, 2020.

Abstract

Learning community structures in graphs that are randomly generated by stochastic block models (SBMs) has received much attention lately. In this paper, we focus on the problem of exactly recovering the communities in a binary symmetric SBM, where a graph of $n$ vertices is partitioned into two equal-sized communities and the vertices are connected with probability $p = \alpha\log(n)/n$ within communities and $q = \beta\log(n)/n$ across communities for some $\alpha>\beta>0$. We propose a two-stage iterative algorithm for solving this problem, which employs the power method with a random starting point in the first-stage and turns to a generalized power method that can identify the communities in a finite number of iterations in the second-stage. It is shown that for any fixed $\alpha$ and $\beta$ such that $\sqrt{\alpha} - \sqrt{\beta} > \sqrt{2}$, which is known to be the information-theoretical limit for exact recovery, the proposed algorithm exactly identifies the underlying communities in $\tilde{O}(n)$ running time with probability tending to one as $n\rightarrow\infty$. We also present numerical results of the proposed algorithm to support and complement our theoretical development.

Cite this Paper


BibTeX
@InProceedings{pmlr-v119-wang20ac, title = {A Nearly-Linear Time Algorithm for Exact Community Recovery in Stochastic Block Model}, author = {Wang, Peng and Zhou, Zirui and So, Anthony Man-Cho}, booktitle = {Proceedings of the 37th International Conference on Machine Learning}, pages = {10126--10135}, year = {2020}, editor = {III, Hal Daumé and Singh, Aarti}, volume = {119}, series = {Proceedings of Machine Learning Research}, month = {13--18 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v119/wang20ac/wang20ac.pdf}, url = {https://proceedings.mlr.press/v119/wang20ac.html}, abstract = {Learning community structures in graphs that are randomly generated by stochastic block models (SBMs) has received much attention lately. In this paper, we focus on the problem of exactly recovering the communities in a binary symmetric SBM, where a graph of $n$ vertices is partitioned into two equal-sized communities and the vertices are connected with probability $p = \alpha\log(n)/n$ within communities and $q = \beta\log(n)/n$ across communities for some $\alpha>\beta>0$. We propose a two-stage iterative algorithm for solving this problem, which employs the power method with a random starting point in the first-stage and turns to a generalized power method that can identify the communities in a finite number of iterations in the second-stage. It is shown that for any fixed $\alpha$ and $\beta$ such that $\sqrt{\alpha} - \sqrt{\beta} > \sqrt{2}$, which is known to be the information-theoretical limit for exact recovery, the proposed algorithm exactly identifies the underlying communities in $\tilde{O}(n)$ running time with probability tending to one as $n\rightarrow\infty$. We also present numerical results of the proposed algorithm to support and complement our theoretical development.} }
Endnote
%0 Conference Paper %T A Nearly-Linear Time Algorithm for Exact Community Recovery in Stochastic Block Model %A Peng Wang %A Zirui Zhou %A Anthony Man-Cho So %B Proceedings of the 37th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2020 %E Hal Daumé III %E Aarti Singh %F pmlr-v119-wang20ac %I PMLR %P 10126--10135 %U https://proceedings.mlr.press/v119/wang20ac.html %V 119 %X Learning community structures in graphs that are randomly generated by stochastic block models (SBMs) has received much attention lately. In this paper, we focus on the problem of exactly recovering the communities in a binary symmetric SBM, where a graph of $n$ vertices is partitioned into two equal-sized communities and the vertices are connected with probability $p = \alpha\log(n)/n$ within communities and $q = \beta\log(n)/n$ across communities for some $\alpha>\beta>0$. We propose a two-stage iterative algorithm for solving this problem, which employs the power method with a random starting point in the first-stage and turns to a generalized power method that can identify the communities in a finite number of iterations in the second-stage. It is shown that for any fixed $\alpha$ and $\beta$ such that $\sqrt{\alpha} - \sqrt{\beta} > \sqrt{2}$, which is known to be the information-theoretical limit for exact recovery, the proposed algorithm exactly identifies the underlying communities in $\tilde{O}(n)$ running time with probability tending to one as $n\rightarrow\infty$. We also present numerical results of the proposed algorithm to support and complement our theoretical development.
APA
Wang, P., Zhou, Z. & So, A.M.. (2020). A Nearly-Linear Time Algorithm for Exact Community Recovery in Stochastic Block Model. Proceedings of the 37th International Conference on Machine Learning, in Proceedings of Machine Learning Research 119:10126-10135 Available from https://proceedings.mlr.press/v119/wang20ac.html.

Related Material