Advertisement

Efficient and Practical Tree Preconditioning for Solving Laplacian Systems

  • Luca Castelli Aleardi
  • Alexandre Nolin
  • Maks Ovsjanikov
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9125)

Abstract

We consider the problem of designing efficient iterative methods for solving linear systems. In its full generality, this is one of the oldest problems in numerical analysis with a tremendous number of practical applications. We focus on a particular type of linear systems, associated with Laplacian matrices of undirected graphs, and study a class of iterative methods for which it is possible to speed up the convergence through combinatorial preconditioning. We consider a class of preconditioners, known as tree preconditioners, introduced by Vaidya, that have been shown to lead to asymptotic speed-up in certain cases. Rather than trying to improve the structure of the trees used in preconditioning, we propose a very simple modification to the basic tree preconditioner, which can significantly improve the performance of the iterative linear solvers in practice. We show that our modification leads to better conditioning for some special graphs, and provide extensive experimental evidence for the decrease in the complexity of the preconditioned conjugate gradient method for several graphs, including 3D meshes and complex networks.

Keywords

Span Tree Condition Number Conjugate Gradient Spectral Cluster Laplacian Matrix 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Notes

Acknowledgments

This work is supported by the ANR EGOS 12 JS02 002 01, a Google Faculty Research Award, the Marie Curie grant CIG-334283-HRGP, and a CNRS chaire dexcellence, Jean Marjoulet professorial chair.

References

  1. 1.
    Abraham, I., Neiman, O.: Using petal-decompositions to build a low stretch spanning tree. In: Proc. STOC, pp. 395–406. ACM (2012)Google Scholar
  2. 2.
    Alon, N., Karp, R.M., Peleg, D., West, D.B.: A graph-theoretic game and its application to the k-server problem. SIAM J. Comput. 24(1), 78–100 (1995)zbMATHMathSciNetCrossRefGoogle Scholar
  3. 3.
    Batson, J.D., Spielman, D.A., Srivastava, N.: Twice-ramanujan sparsifiers. SIAM Review 56(2), 315–334 (2014)zbMATHMathSciNetCrossRefGoogle Scholar
  4. 4.
    Batson, J.D., Spielman, D.A., Srivastava, N., Teng, S.: Spectral sparsification of graphs: theory and algorithms. Commun. ACM 56(8), 87–94 (2013)CrossRefGoogle Scholar
  5. 5.
    Beauwens, R.: Lower eigenvalue bounds for pencils of matrices. Linear Algebra and its Applications 85, 101–119 (1987)zbMATHMathSciNetCrossRefGoogle Scholar
  6. 6.
    Bern, M.W., Gilbert, J.R., Hendrickson, B., Nguyen, N., Toledo, S.: Support-graph preconditioners. SIAM J. Matrix Analysis Applications 27(4), 930–951 (2006)zbMATHMathSciNetCrossRefGoogle Scholar
  7. 7.
    Boman, E., Hendrickson, B.: Support theory for preconditioning. SIAM Journal on Matrix Analysis and Applications 25(3), 694–717 (2003)zbMATHMathSciNetCrossRefGoogle Scholar
  8. 8.
    Castelli Aleardi, L.C., Nolin, A., Ovsjanikov, M.: Efficient and practical tree preconditioning for solving Laplacian systems (2015). Preprint https://hal.inria.fr/hal-01138603
  9. 9.
    Chen, D., Toledo, S.: Vaidya’s preconditioners: implementation and experimental study. Elect. Trans. on Numerical Analysis 16, 30–49 (2003)zbMATHMathSciNetGoogle Scholar
  10. 10.
    Cohen, M.B., Kyng, R., Miller, G.L., Pachocki, J., Peng, R., Rao, A., Xu, S.C.: Solving SDD linear systems in nearly \(m\log ^{1/2}\)n time. STOC, pp. 343–352 (2014)Google Scholar
  11. 11.
    Cohen, M.B., Miller, G.L., Pachocki, J.W., Peng, R., Xu, S.C.: Stretching stretch (2014). CoRR, abs/1401.2454Google Scholar
  12. 12.
    Golub, G.H., Van Loan, C.F.: Matrix Computations. 4th edn (2013)Google Scholar
  13. 13.
    Hoske, D., Lukarski, D., Meyerhenke, H., Wegner, M.: Is nearly-linear the same in theory and practice? a case study with a combinatorial laplacian solver. In: Proc, SEA (2015)Google Scholar
  14. 14.
    Kolla, A., Makarychev, Y., Saberi, A., Teng, S.-H.: Subgraph sparsification and nearly optimal ultrasparsifiers. In: Proc. STOC, pp. 57–66. ACM (2010)Google Scholar
  15. 15.
    Koren, Y.: Drawing graphs by eigenvectors: Theory and practice. Computers and Mathematics with Applications 49, 2005 (2005)MathSciNetCrossRefGoogle Scholar
  16. 16.
    Koutis, I., Miller, G.L., Peng, R.: A fast solver for a class of linear systems. Commun. ACM 55(10), 99–107 (2012)CrossRefGoogle Scholar
  17. 17.
    Krishnan, D., Fattal, R., Szeliski, R.: Efficient preconditioning of Laplacian matrices for Computer Graphics. ACM Trans. Graph. 32(4), 142 (2013)CrossRefGoogle Scholar
  18. 18.
    Poulalhon, D., Schaeffer, G.: Optimal coding and sampling of triangulations. Algorithmica 46(3–4), 505–527 (2006)zbMATHMathSciNetCrossRefGoogle Scholar
  19. 19.
    Spielman, D.A., Teng, S.: Spectral sparsification of graphs. SIAM J. Comput. 40(4), 981–1025 (2011)zbMATHMathSciNetCrossRefGoogle Scholar
  20. 20.
    Spielman, D.A., Teng, S.: A local clustering algorithm for massive graphs and its application to nearly linear time graph partitioning. SIAM J. Comput. 42(1), 1–26 (2013)zbMATHMathSciNetCrossRefGoogle Scholar
  21. 21.
    Spielman, D.A., Teng, S.: Nearly linear time algorithms for preconditioning and solving symmetric, diagonally dominant linear systems. SIAM J. Matrix Analysis Applications 35(3), 835–885 (2014)zbMATHMathSciNetCrossRefGoogle Scholar
  22. 22.
    Spielman, D.A., Teng, S.-H.: Nearly-linear time algorithms for graph partitioning, graph sparsification, and solving linear systems. STOC, pp. 81–90 (2004)Google Scholar
  23. 23.
    Vaidya, P.M.: Solving linear equations with symmetric diagonally dominant matrices by constructing good preconditioners. Unpublished manuscript (1991)Google Scholar
  24. 24.
    Vishnoi, N.K.: Lx = b. Foundations and Trends in Theoretical Computer Science 8(1–2), 1–141 (2013)MathSciNetCrossRefGoogle Scholar
  25. 25.
    von Luxburg, U.: A tutorial on spectral clustering. Statistics and Computing 17(4), 395–416 (2007)MathSciNetCrossRefGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  • Luca Castelli Aleardi
    • 1
  • Alexandre Nolin
    • 1
  • Maks Ovsjanikov
    • 1
  1. 1.LIX - École PolytechniquePalaiseauFrance

Personalised recommendations