Advanced search
1 file | 251.71 KB

Initialization dependence of clustering algorithms

Author
Organization
Abstract
It is well known that the clusters produced by a clustering algorithm depend on the chosen initial centers. In this paper we present a measure for the degree to which a given clustering algorithm depends on the choice of initial centers, for a given data set. This measure is calculated for four well-known offline clustering algorithms (k-means Forgy, k-means Hartigan, k-means Lloyd and frizzy c-means), for five benchmark data sets. The measure is also calculated for ECM, an online algorithm that does not require the number of initial centers as input, but for which the resulting clusters can depend oil the order that the input arrives. Our main finding is that this initialization dependence measure call also he used to determine the optimal number of clusters.

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 251.71 KB

Citation

Please use this url to cite or link to this publication:

Chicago
De Mulder, Wim, Stefan Schliebs, René Boel, and Martin Kuiper. 2009. “Initialization Dependence of Clustering Algorithms.” In Lecture Notes in Computer Science, ed. Mario Köppen, Nikola Kasabov, and George Coghill, 5507:615–622. Berlin, Germany: Springer.
APA
De Mulder, W., Schliebs, S., Boel, R., & Kuiper, M. (2009). Initialization dependence of clustering algorithms. In M. Köppen, N. Kasabov, & G. Coghill (Eds.), LECTURE NOTES IN COMPUTER SCIENCE (Vol. 5507, pp. 615–622). Presented at the 15th International Conference on Neuro-Information Processing, Berlin, Germany: Springer.
Vancouver
1.
De Mulder W, Schliebs S, Boel R, Kuiper M. Initialization dependence of clustering algorithms. In: Köppen M, Kasabov N, Coghill G, editors. LECTURE NOTES IN COMPUTER SCIENCE. Berlin, Germany: Springer; 2009. p. 615–22.
MLA
De Mulder, Wim et al. “Initialization Dependence of Clustering Algorithms.” Lecture Notes in Computer Science. Ed. Mario Köppen, Nikola Kasabov, & George Coghill. Vol. 5507. Berlin, Germany: Springer, 2009. 615–622. Print.
@inproceedings{790422,
  abstract     = {It is well known that the clusters produced by a clustering algorithm depend on the chosen initial centers. In this paper we present a measure for the degree to which a given clustering algorithm depends on the choice of initial centers, for a given data set. This measure is calculated for four well-known offline clustering algorithms (k-means Forgy, k-means Hartigan, k-means Lloyd and frizzy c-means), for five benchmark data sets. The measure is also calculated for ECM, an online algorithm that does not require the number of initial centers as input, but for which the resulting clusters can depend oil the order that the input arrives. Our main finding is that this initialization dependence measure call also he used to determine the optimal number of clusters.},
  author       = {De Mulder, Wim and Schliebs, Stefan and Boel, Ren{\'e} and Kuiper, Martin},
  booktitle    = {LECTURE NOTES IN COMPUTER SCIENCE},
  editor       = {K{\"o}ppen, Mario and Kasabov, Nikola and Coghill, George},
  isbn         = {978-3-642-03039-0},
  issn         = {0302-9743},
  language     = {eng},
  location     = {Auckland, New Zealand},
  pages        = {615--622},
  publisher    = {Springer},
  title        = {Initialization dependence of clustering algorithms},
  url          = {http://dx.doi.org/10.1007/978-3-642-03040-6\_75},
  volume       = {5507},
  year         = {2009},
}

Altmetric
View in Altmetric
Web of Science
Times cited: