ACM Home Page
Please provide us with feedback. Feedback
A decentralized algorithm for spectral analysis
Full text PdfPdf (167 KB)
Source
Annual ACM Symposium on Theory of Computing archive
Proceedings of the thirty-sixth annual ACM symposium on Theory of computing table of contents
Chicago, IL, USA
SESSION: Session 15B table of contents
Pages: 561 - 568  
Year of Publication: 2004
ISBN:1-58113-852-0
Authors
David Kempe  University of Washington
Frank McSherry  Microsoft Research SVC
Sponsors
ACM: Association for Computing Machinery
SIGACT: ACM Special Interest Group on Algorithms and Computation Theory
Publisher
ACM  New York, NY, USA
Bibliometrics
Downloads (6 Weeks): 17,   Downloads (12 Months): 121,   Citation Count: 9
Additional Information:

abstract   references   cited by   index terms   review   collaborative colleagues  

Tools and Actions: Request Permissions Request Permissions    Review this Article  
DOI Bookmark: Use this link to bookmark this Article: http://doi.acm.org/10.1145/1007352.1007438
What is a DOI?

ABSTRACT

In many large network settings, such as computer networks, social networks, or hyperlinked text documents, much information can be obtained from the network's spectral properties. However, traditional centralized approaches for computing eigenvectors struggle with at least two obstacles: the data may be difficult to obtain (both due to technical reasons and because of privacy concerns), and the sheer size of the networks makes the computation expensive. A decentralized, distributed algorithm addresses both of these obstacles: it utilizes the computational power of all nodes in the network and their ability to communicate, thus speeding up the computation with the network size. And as each node knows its incident edges, the data collection problem is avoided as well.Our main result is a simple decentralized algorithm for computing the top k eigenvectors of a symmetric weighted adjacency matrix, and a proof that it converges essentially in OMIXlog2 n) rounds of communication and computation, where τMIX is the mixing time of a random walk on the network. An additional contribution of our work is a decentralized way of actually detecting convergence, and diagnosing the current error. Our protocol scales well, in that the amount of computation performed at any node in any one round, and the sizes of messages sent, depend polynomially on k, but not on the (typically much larger) number n of nodes.


REFERENCES

Note: OCR errors may be found in this Reference List extracted from the full text article. ACM has opted to expose the complete List rather than only correct and linked references.

 
1
2
 
3
 
4
 
5
S. Boyd, A. Ghosh, B. Prabhakar, and D. Shah. Gossip and mixing times of random walks on random graphs. Submitted.
 
6
 
7
F. Chung. Spectral Graph Theory. American Mathematical Society, 1997.
 
8
S. Deerwester, S. Dumais, T. Landauer, G. Furnas, and R. Harshman. Indexing by latent semantic analysis. J. of the American Society for Information Sciences, 41:391--407, 1990.
 
9
M. Fiedler. A property of eigenvectors of nonnegative symmetric matrices and its applications to graph theory. Czechoslovak Mathematical Journal, 25:619--633, 1975.
 
10
 
11
G. Golub and C. V. Loan. Matrix Computations. Johns Hopkins University Press, third edition, 1996.
 
12
M. Granovetter. The strength of weak ties. American Journal of Sociology, 78:1360--1380, 1973.
 
13
 
14
15
16
 
17
 
18
A. Ng, M. Jordan, and Y. Weiss. On spectral clustering: Analysis and an algorithm. In Proc. 14th Advances in Neural Information Processing Systems, 2002.
19
 
20
 
21
L. Saloff-Coste. Lectures on finite markov chains. In Lecture Notes in Mathematics 1665, pages 301--408. Springer, 1997. École d'été de St. Flour 1996.
 
22
G. Stewart. On the perturbation of LU and cholesky factors. IMA Journal of Numerical Analysis, 1997.
23
 
24
P. Wedin. Perturbation theory for pseudo-inverses. BIT, 13:217--232, 1973.
 
25

CITED BY  9


REVIEW

"Charles Raymond Crawford : Reviewer"

A decentralized algorithm for the computation of the spectral decomposition of a graph is presented in this paper. The algorithm is based on the use of subspace iteration to compute a few dominant eigenvectors of the adjacency matrix of the graph   more...

Collaborative Colleagues:
David Kempe: colleagues
Frank McSherry: colleagues