Clustering with or without the Approximation

  • Frans Schalekamp
  • Michael Yu
  • Anke van Zuylen
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6196)


We study algorithms for clustering data that were recently proposed by Balcan, Blum and Gupta in SODA’09 [4] and that have already given rise to two follow-up papers. The input for the clustering problem consists of points in a metric space and a number k, specifying the desired number of clusters. The algorithms find a clustering that is provably close to a target clustering, provided that the instance has the “( 1 + α, ε)-property”, which means that the instance is such that all solutions to the k-median problem for which the objective value is at most (1 + α) times the optimal objective value correspond to clusterings that misclassify at most an ε fraction of the points with respect to the target clustering. We investigate the theoretical and practical implications of their results.

Our main contributions are as follows. First, we show that instances that have the ( 1 + α, ε)-property and for which, additionally, the clusters in the target clustering are large, are easier than general instances: the algorithm proposed in [4] is a constant factor approximation algorithm with an approximation guarantee that is better than the known hardness of approximation for general instances. Further, we show that it is NP-hard to check if an instance satisfies the ( 1 + α, ε)-property for a given (α, ε); the algorithms in [4] need such α and ε as input parameters, however. We propose ways to use their algorithms even if we do not know values of α and ε for which the assumption holds. Finally, we implement these methods and other popular methods, and test them on real world data sets. We find that on these data sets there are no α and ε so that the dataset has both ( 1 + α, ε)-property and sufficiently large clusters in the target solution. For the general case, we show that on our data sets the performance guarantee proved by [4] is meaningless for the values of α, ε such that the data set has the ( 1 + α, ε)-property. The algorithm nonetheless gives reasonable results, although it is outperformed by other methods.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Arthur, D., Vassilvitskii, S.: k-means++: the advantages of careful seeding. In: SODA ’07: 18th Annual ACM-SIAM Symposium on Discrete Algorithms, pp. 1027–1035 (2007)Google Scholar
  2. 2.
    Arya, V., Garg, N., Khandekar, R., Meyerson, A., Munagala, K., Pandit, V.: Local search heuristics for k-median and facility location problems. SIAM J. Comput. 33(3), 544–562 (2004)zbMATHCrossRefMathSciNetGoogle Scholar
  3. 3.
    Asuncion, A., Newman, D.: UCI machine learning repository (2007),
  4. 4.
    Balcan, M.-F., Blum, A., Gupta, A.: Approximate clustering without the approximation. In: SODA ’09: 19th Annual ACM -SIAM Symposium on Discrete Algorithms, pp. 1068–1077 (2009)Google Scholar
  5. 5.
    Balcan, M.-F., Blum, A., Vempala, S.: A discriminative framework for clustering via similarity functions. In: STOC 2008: 40th Annual ACM Symposium on Theory of Computing, pp. 671–680 (2008)Google Scholar
  6. 6.
    Balcan, M.-F., Braverman, M.: Finding low error clusterings. In: COLT 2009: 22nd Annual Conference on Learning Theory (2009)Google Scholar
  7. 7.
    Balcan, M.-F., Röglin, H., Teng, S.-H.: Agnostic clustering. In: Gavaldà, R., Lugosi, G., Zeugmann, T., Zilles, S. (eds.) ALT 2009. LNCS, vol. 5809, pp. 384–398. Springer, Heidelberg (2009)CrossRefGoogle Scholar
  8. 8.
    Beasley, J.E.: A note on solving large p-median problems. European Journal of Operational Research 21(2), 270–273 (1985)zbMATHCrossRefMathSciNetGoogle Scholar
  9. 9.
    Beasley, J.E.: OR-Library p-median - uncapacitated (1985),
  10. 10.
    Bilu, Y., Linial, N.: Are stable instances easy. In: ICS 2010: The First Symposium on Innovations in Computer Science, pp. 332–341 (2010)Google Scholar
  11. 11.
    Gupta, A.: Personal Communication (2009)Google Scholar
  12. 12.
    Jain, K., Mahdian, M., Markakis, E., Saberi, A., Vazirani, V.V.: Greedy facility location algorithms analyzed using dual fitting with factor-revealing LP. J. ACM 50(6), 795–824 (2003)CrossRefMathSciNetGoogle Scholar
  13. 13.
    Jain, K., Vazirani, V.V.: Approximation algorithms for metric facility location and k-median problems using the primal-dual schema and Lagrangian relaxation. J. ACM 48(2), 274–296 (2001)zbMATHCrossRefMathSciNetGoogle Scholar
  14. 14.
    Lloyd, S.: Least squares quantization in PCM. IEEE Transactions on Information Theory 28(2), 129–137 (1982)zbMATHCrossRefMathSciNetGoogle Scholar
  15. 15.
    Ostrovsky, R., Rabani, Y., Schulman, L.J., Swamy, C.: The effectiveness of Lloyd-type methods for the k-means problem. In: FOCS ’06:47th Annual IEEE Symposium on Foundations of Computer Science, pp. 165–176 (2006)Google Scholar
  16. 16.
    Schalekamp, F., Yu, M., van Zuylen, A.: Clustering with or without the approximation,

Copyright information

© Springer-Verlag Berlin Heidelberg 2010

Authors and Affiliations

  • Frans Schalekamp
    • 1
  • Michael Yu
    • 2
  • Anke van Zuylen
    • 1
  1. 1.ITCSTsinghua University 
  2. 2.MIT 

Personalised recommendations