Large scale spectral clustering using resistance distance and spielman-teng solvers

Nguyen Lu Dang Khoa, Sanjay Chawla

Research output: Chapter in Book/Report/Conference proceedingConference contribution

14 Citations (Scopus)

Abstract

The promise of spectral clustering is that it can help detect complex shapes and intrinsic manifold structure in large and high dimensional spaces. The price for this promise is the computational cost O(n 3) for computing the eigen-decomposition of the graph Laplacian matrix-so far a necessary subroutine for spectral clustering. In this paper we bypass the eigen-decomposition of the original Laplacian matrix by leveraging the recently introduced Spielman and Teng near-linear time solver for systems of linear equations and random projection. Experiments on several synthetic and real datasets show that the proposed approach has better clustering quality and is faster than the state-of-the-art approximate spectral clustering methods.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages7-21
Number of pages15
Volume7569 LNAI
DOIs
Publication statusPublished - 2012
Externally publishedYes
Event15th International Conference on Discovery Science, DS 2012 - Lyon
Duration: 29 Oct 201231 Oct 2012

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7569 LNAI
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other15th International Conference on Discovery Science, DS 2012
CityLyon
Period29/10/1231/10/12

Fingerprint

Spectral Clustering
Laplacian Matrix
Decomposition
Subroutines
Linear equations
Random Projection
Graph Laplacian
Decompose
Spectral Methods
System of Linear Equations
Clustering Methods
Linear Time
Computational Cost
High-dimensional
Clustering
Costs
Necessary
Computing
Experiments
Experiment

Keywords

  • random projection
  • resistance distance
  • spectral clustering
  • Spielman-Teng Solver

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Khoa, N. L. D., & Chawla, S. (2012). Large scale spectral clustering using resistance distance and spielman-teng solvers. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7569 LNAI, pp. 7-21). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7569 LNAI). https://doi.org/10.1007/978-3-642-33492-4_4

Large scale spectral clustering using resistance distance and spielman-teng solvers. / Khoa, Nguyen Lu Dang; Chawla, Sanjay.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7569 LNAI 2012. p. 7-21 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7569 LNAI).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Khoa, NLD & Chawla, S 2012, Large scale spectral clustering using resistance distance and spielman-teng solvers. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 7569 LNAI, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7569 LNAI, pp. 7-21, 15th International Conference on Discovery Science, DS 2012, Lyon, 29/10/12. https://doi.org/10.1007/978-3-642-33492-4_4
Khoa NLD, Chawla S. Large scale spectral clustering using resistance distance and spielman-teng solvers. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7569 LNAI. 2012. p. 7-21. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-33492-4_4
Khoa, Nguyen Lu Dang ; Chawla, Sanjay. / Large scale spectral clustering using resistance distance and spielman-teng solvers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7569 LNAI 2012. pp. 7-21 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{903ab985e62c4a26935b53777d384d4f,
title = "Large scale spectral clustering using resistance distance and spielman-teng solvers",
abstract = "The promise of spectral clustering is that it can help detect complex shapes and intrinsic manifold structure in large and high dimensional spaces. The price for this promise is the computational cost O(n 3) for computing the eigen-decomposition of the graph Laplacian matrix-so far a necessary subroutine for spectral clustering. In this paper we bypass the eigen-decomposition of the original Laplacian matrix by leveraging the recently introduced Spielman and Teng near-linear time solver for systems of linear equations and random projection. Experiments on several synthetic and real datasets show that the proposed approach has better clustering quality and is faster than the state-of-the-art approximate spectral clustering methods.",
keywords = "random projection, resistance distance, spectral clustering, Spielman-Teng Solver",
author = "Khoa, {Nguyen Lu Dang} and Sanjay Chawla",
year = "2012",
doi = "10.1007/978-3-642-33492-4_4",
language = "English",
isbn = "9783642334917",
volume = "7569 LNAI",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "7--21",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Large scale spectral clustering using resistance distance and spielman-teng solvers

AU - Khoa, Nguyen Lu Dang

AU - Chawla, Sanjay

PY - 2012

Y1 - 2012

N2 - The promise of spectral clustering is that it can help detect complex shapes and intrinsic manifold structure in large and high dimensional spaces. The price for this promise is the computational cost O(n 3) for computing the eigen-decomposition of the graph Laplacian matrix-so far a necessary subroutine for spectral clustering. In this paper we bypass the eigen-decomposition of the original Laplacian matrix by leveraging the recently introduced Spielman and Teng near-linear time solver for systems of linear equations and random projection. Experiments on several synthetic and real datasets show that the proposed approach has better clustering quality and is faster than the state-of-the-art approximate spectral clustering methods.

AB - The promise of spectral clustering is that it can help detect complex shapes and intrinsic manifold structure in large and high dimensional spaces. The price for this promise is the computational cost O(n 3) for computing the eigen-decomposition of the graph Laplacian matrix-so far a necessary subroutine for spectral clustering. In this paper we bypass the eigen-decomposition of the original Laplacian matrix by leveraging the recently introduced Spielman and Teng near-linear time solver for systems of linear equations and random projection. Experiments on several synthetic and real datasets show that the proposed approach has better clustering quality and is faster than the state-of-the-art approximate spectral clustering methods.

KW - random projection

KW - resistance distance

KW - spectral clustering

KW - Spielman-Teng Solver

UR - http://www.scopus.com/inward/record.url?scp=84868113132&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84868113132&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-33492-4_4

DO - 10.1007/978-3-642-33492-4_4

M3 - Conference contribution

SN - 9783642334917

VL - 7569 LNAI

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 7

EP - 21

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -