Abstract
Spectral partitioning is a simple, nearly linear time algorithm to find sparse cuts, and the Cheeger inequalities provide a worst-case guarantee for the quality of the approximation found by the algorithm. A local graph partitioning algorithm finds a set of vertices with small conductance (i.e., a sparse cut) by adaptively exploring part of a large graph G, starting from a specified vertex. For the algorithm to be local, its complexity must be bounded in terms of the size of the set that it outputs, with at most a weak dependence on the number n of vertices in G. Previous local partitioning algorithms find sparse cuts using random walks and personalized PageRank [Spielman and Teng 2013; Andersen et al. 2006].
In this article, we introduce a simple randomized local partitioning algorithm that finds a sparse cut by simulating the volume-biased evolving set process, which is a Markov chain on sets of vertices. We prove that for any ϵ > 0, and any set of vertices A that has conductance at most φ, for at least half of the starting vertices in A our algorithm will output (with constant probability) a set of conductance O(√φ /ϵ). We prove that for a given run of the algorithm, the expected ratio between its computational complexity and the volume of the set that it outputs is vol(A)ϵφ-1/2polylog(n), where vol(A) = Σv ∈ Ad(v) is the volume of the set A. This gives an algorithm with the same guarantee (up to a constant factor) as the Cheeger's inequality that runs in time slightly superlinear in the size of the output. This is the first sublinear (in the size of the input) time algorithm with almost the same guarantee as the Cheeger's inequality. In comparison, the best previous local partitioning algorithm, by Andersen et al. [2006], has a worse approximation guarantee of O(√φ log n) and a larger ratio of φ-1 polylog(n) between the complexity and output volume.
As a by-product of our results, we prove a bicriteria approximation algorithm for the expansion profile of any graph. For 0 < k ≤ vol(V)/2, let φ(k) : min S: vol(S) ≤ kφ(S). There is a polynomial time algorithm that, for any k, ϵ > 0, finds a set S of volume vol(S) ≤ O(k1 + ϵ) and expansion φ(S)≤ O(√φ (k)/ϵ). As a new technical tool, we show that for any set S of vertices of a graph, a lazy t-step random walk started from a randomly chosen vertex of S will remain entirely inside S with probability at least (1 - φ(S)/2)t. This itself provides a new lower bound to the uniform mixing time of any finite state reversible Markov chain.
- N Alon. 1986. Eigen values and expanders. Combinatorica 6 (January 1986), 83--96. Issue 2. http://portal.acm.org/citation.cfm?id=18497.18498. Google ScholarDigital Library
- N. Alon and V. Milman. 1985. Isoperimetric inequalities for graphs, and superconcentrators. J. Combin. Theor. Ser. B 38, 1 (Feb 1985), 73--88.Google ScholarCross Ref
- Reid Andersen, Fan R. K. Chung, and Kevin J. Lang. 2006. Local graph partitioning using pagerank vectors. In FOCS. 475--486. Google ScholarDigital Library
- Reid Andersen and Kevin J. Lang. 2006. Communities from seed sets. In WWW. ACM, New York, NY, 223--232. Google ScholarDigital Library
- Reid Andersen and Yuval Peres. 2009. Finding sparse cuts locally using evolving sets. In STOC. 235--244. Google ScholarDigital Library
- Sanjeev Arora, Boaz Barak, and David Steurer. 2010a. Subexponential algorithms for unique games and related problems. In FOCS. IEEE Computer Society, Washington, DC, 563--572. Google ScholarDigital Library
- Sanjeev Arora, Elad Hazan, and Satyen Kale. 2010b. O(&sqrt;log n) Approximation to sparsest cut in Õ(n<sup>2</sup>) time. SIAM J. Comput. 39, 5 (2010), 1748--1771.Google ScholarDigital Library
- Nikhil Bansal, Uriel Feige, Robert Krauthgamer, Konstantin Makarychev, Viswanath Nagarajan, Joseph Naor, and Roy Schwartz. 2011. Min-max graph partitioning and small set expansion. In FOCS. IEEE, Washington, DC, 17--26. Google ScholarDigital Library
- G. R. Blakley and Prabir Roy. 1965. A hölder type inequality for symmetric matrices with nonnegative entries. In Proc. Am. Math. Soc. Vol. 16. 1244--1245.Google Scholar
- Siu On Chan, Tsz Chiu Kwok, and Lap Chi Lau. 2015. Random Walks and Evolving Sets: Faster Convergences and Limitations. (2015). http://arxiv.org/abs/1507.02069.Google Scholar
- T. H. Cormen, C. E. Leiserson, R. L. Rivest, and C. Stein. 2009. Introduction to Algorithms (3rd ed.). MIT Press, Cambridge, MA. Google ScholarDigital Library
- Persi Diaconis and James A. Fill. 1990. Strong stationary times via a new form of duality. Ann. Probab. 18, 4 (1990), 1483--1522.Google ScholarCross Ref
- Gary William Flake, Steve Lawrence, and C. Lee Giles. 2000. Efficient identification of Web communities. In SIGKDD. ACM, New York, NY, 150--160. Google ScholarDigital Library
- M. Jerrum and Alistair Sinclair. 1989. Approximating the permanent. SIAM J. Comput. 18, 6 (1989), 1149--1178. Google ScholarDigital Library
- Ravi Kannan, Santosh Vempala, and Adrian Vetta. 2004. On clusterings: Good, bad and spectral. J. ACM 51 (2004), 497--515. Google ScholarDigital Library
- Jon M. Kleinberg. 1999. Authoritative sources in a hyperlinked environment. J. ACM 46 (1999), 668--677. Google ScholarDigital Library
- Tsz Chiu Kwok and Lap Chi Lau. 2012. Finding small sparse cuts by random walk. In APPROX-RANDOM. 615--626.Google Scholar
- Tsz Chiu Kwok, Lap Chi Lau, Yin Tat Lee, Shayan Oveis Gharan, and Luca Trevisan. 2013. Improved cheeger’s inequality: Analysis of spectral partitioning algorithms through higher order spectral gap. In STOC. 11--20. Google ScholarDigital Library
- Jure Leskovec, Kevin J. Lang, Anirban Dasgupta, and Michael W. Mahoney. 2008. Statistical properties of community structure in large social and information networks. In WWW. ACM, New York, NY, 695--704. Google ScholarDigital Library
- Jure Leskovec, Kevin J. Lang, Anirban Dasgupta, and Michael W. Mahoney. 2009. Community structure in large networks: Natural cluster sizes and the absence of large well-defined clusters. Internet Math. 6, 1 (2009), 29--123.Google ScholarCross Ref
- Jure Leskovec, Kevin J. Lang, and Michael Mahoney. 2010. Empirical comparison of algorithms for network community detection. In WWW. New York, NY, 631--640. Google ScholarDigital Library
- David A. Levin, Yuval Peres, and Elizabeth L. Wilmer. 2006. Markov chains and mixing times. American Mathematical Society.Google Scholar
- Angsheng Li, Yicheng Pan, and Pan Peng. 2011. Testing conductance in general graphs. Electron. Colloq. Comput. Complex. (ECCC) 18 (2011), 101.Google Scholar
- Angsheng Li and Pan Peng. 2011. Community structures in classical network models. Internet Math. 7, 2 (2011), 81--106.Google ScholarCross Ref
- David London. 1966. Inequalities in quadratic forms. Duke Math. J. 33, 3 (1966), 511--522.Google ScholarCross Ref
- László Lovász and Ravi Kannan. 1999. Faster mixing via average conductance. In STOC. ACM, 282--287. Google ScholarDigital Library
- Ravi Montenegro. 2007. Sharp edge, vertex, and mixed cheeger inequalities for finite markov kernels. Electronic Communications in Probability {electronic only} 12 (2007), 377--389.Google Scholar
- Ravi Montenegro. 2009. The simple random walk and max-degree walk on a directed graph. Rand. Struct. Algorithm. 34, 3 (May 2009), 395--407. Google ScholarDigital Library
- Ben Morris and Yuval Peres. 2003. Evolving sets and mixin. In STOC. 279--286. Google ScholarDigital Library
- H. P. Mulholland and C. A. B. Smith. 1959. An inequality arising in genetical theory. Am. Math. Month. 66 (1959), 673--683.Google ScholarCross Ref
- Ryan O’Donnell and David Witmer. 2012. Improved small-set expansion from higher eigenvalues. CoRR abs/1204.4688 (2012).Google Scholar
- Lorenzo Orecchia, Sushant Sachdeva, and Nisheeth K. Vishnoi. 2012. Approximating the exponential, the lanczos method and an Õ(m)-time spectral algorithm for balanced separator. In STOC. Google ScholarDigital Library
- Lorenzo Orecchia and Nisheeth K. Vishnoi. 2011. Towards an sdp-based approach to spectral methods: A nearly-linear-time algorithm for graph partitioning and decomposition. In SODA. 532--545. Google ScholarDigital Library
- Shayan Oveis Gharan and Luca Trevisan. 2012. Approximating the expansion profile and almost optimal local graph clustering. In FOCS. 187--196. Google ScholarDigital Library
- Prasad Raghavendra and David Steurer. 2010. Graph expansion and the unique games conjecture. In STOC. ACM, New York, NY, 755--764. Google ScholarDigital Library
- Prasad Raghavendra, David Steurer, and Prasad Tetali. 2010. Approximations for the isoperimetric and spectral profile of graphs and related parameters. In STOC. ACM, New York, NY, 631--640. Google ScholarDigital Library
- Jonah Sherman. 2009. Breaking the multicommodity flow barrier for O(&sqrt;log n)-approximations to sparsest cut. In FOCS. 363--372. Google ScholarDigital Library
- Jianbo Shi and Jitendra Malik. 2000. Normalized cuts and image segmentation. IEEE Trans. Pattern Anal. Mach. Intell. 22, 8 (2000), 888--905. Google ScholarDigital Library
- Daniel A. Spielman and Shang-Hua Teng. 2004. Nearly-linear time algorithms for graph partitioning, graph sparsification, and solving linear systems. In STOC. 81--90. Google ScholarDigital Library
- Daniel A. Spielman and Shang-Hua Teng. 2013. A local clustering algorithm for massive graphs and its application to nearly linear time graph partitioning. SIAM J. Comput. 42, 1 (2013), 1--26.Google ScholarDigital Library
- David A. Tolliver and Gary L. Miller. 2006. Graph partitioning by spectral rounding: Applications in image segmentation and clustering. In CVPR. IEEE Computer Society, Washington, DC, 1053--1060. Google ScholarDigital Library
- David Williams. 1991. Probability with Martingales (Cambridge Mathematical Textbooks). Cambridge University Press, Cambridge.Google Scholar
Index Terms
- Almost Optimal Local Graph Clustering Using Evolving Sets
Recommendations
Multiway Spectral Partitioning and Higher-Order Cheeger Inequalities
A basic fact in spectral graph theory is that the number of connected components in an undirected graph is equal to the multiplicity of the eigenvalue zero in the Laplacian matrix of the graph. In particular, the graph is disconnected if and only if ...
Approximating the Expansion Profile and Almost Optimal Local Graph Clustering
FOCS '12: Proceedings of the 2012 IEEE 53rd Annual Symposium on Foundations of Computer ScienceSpectral partitioning is a simple, nearly-linear time, algorithm to find sparse cuts, and the Cheeger inequalities provide a worst-case guarantee of the quality of the approximation found by the algorithm. Local graph partitioning algorithms [ST08, ...
Near-optimal distortion bounds for embedding doubling spaces into L1
STOC '11: Proceedings of the forty-third annual ACM symposium on Theory of computingWe exhibit an infinite doubling metric space (X,d) such that for any non-expansive f : X -> L1, there exists a pair x,y ∈ X with d(x,y) arbitrarily large, and such that |f(x)-f(y)\|1/d(x,y) ≲ √log log d(x,y)}/(log d(x,y)).
As a consequence, we show that ...
Comments