Advertisement

K-Strange Points Clustering Algorithm

Conference paper
Part of the Smart Innovation, Systems and Technologies book series (SIST, volume 31)

Abstract

The classical K-Means clustering algorithm yields means which can be called the final unchanging or fixed means around which all other points in the dataset get clustered. This is so because the K-Means clustering terminates when either the clusters repeat in the next iteration or when the means repeat in the next iteration. This reveals that if one is able to somehow calculate and find apriori the final unchanging means using the dataset, then the task of clustering reduces to only assigning the remaining points in the dataset into clusters, which are closest to these final fixed or unchanging means based on standard distance measures. Taking a cue from the result of the classical K-Means method, the K-Strange points clustering algorithm presented in this paper locates K points from the dataset equaling the number of required clusters which are farthest from each other and are hence called K-Strange points based on the Euclidean distance measure. The remaining points in the dataset are assigned to clusters formed by these K-Strange points.

Keywords

K-Strange points clustering Farthest points Euclidean distance measure 

References

  1. 1.
    Abbas, O.: Comparisons between data clustering algorithms. Int. Arab J. Inf. Technol. 5(3), 320–325 (2008)Google Scholar
  2. 2.
    Prabhu, P., Anbazhagan, N.: Improving the performance of k-means clustering for high dimensional dataset. Int. J. Comput. Sci. Eng. 3(6), 2317–2322 (2011), ISSN: 0975-3397Google Scholar
  3. 3.
    Micheal, J.A.: Berry Gordon Linoff.: Mastering Data Mining. Wiley, Singapore (2001)Google Scholar
  4. 4.
    Bouveyrona, C., Girarda, S., Schmid, C.: High dimensional data clustering. J. Comput. Stat. Data Anal. 52(1), 502–519 (2007)CrossRefGoogle Scholar
  5. 5.
    Kaufman, L., Rousseeuw, P.J.: Finding Groups in Data: An Introduction to Cluster Analysis. Wiley, New York (1990)CrossRefGoogle Scholar
  6. 6.
    Jain, A., Murty, M., Flynn, P.: Data clustering: a review. ACM Comput. Surv. 31(3), 264–323 (1999)CrossRefGoogle Scholar
  7. 7.
    Alijammaat, A., Khalilian, M., Mustapha, N.: A novel approach for high dimensional data clustering. In: Proceedings of the Third International Conference on Knowledge Discovery and Data Mining, Phuket, Iran, pp. 264–267 (2010)Google Scholar
  8. 8.
    Johnson, T.: Bisecting collinear clustering algorithm. Int. J. Comput. Sci. Eng. Inf. Technol. Res. 3(5), 43–46 (2013), © TJPRC Pvt. Ltd., ISSN: 2249-6831Google Scholar
  9. 9.
    Johnson. T., Lobo, J.Z.: Collinear clustering algorithm in lower dimensions. IOSR J. Comput. Eng. 6(5), 08–11 (2012), ISSN: 2278-0661, ISBN: 2278-8727Google Scholar
  10. 10.
    Singh, S.K., Johnson, T.: Improved collinear clustering algorithm in lower dimensions. In: Proceedings of Second International Conference on Emerging Research in Computing, Information, Communication and Applications (2014) (in press)Google Scholar
  11. 11.
    Nagi, S,. Bhattacharya, D.K., Kalita, J.K.: A preview on subspace clustering of high dimensional data. Int. J. Comput. Technol., 6(3), 441–448 (2013). ISSN: 22773061Google Scholar
  12. 12.
    Aravinder D.J., Naganathan, E.R.: Efficient centroids based clustering algorithm with data intelligence. J. Theor. Appl. Inf. Technol. 56(1), 126–130 (2013). ISSN: 1992-8645Google Scholar
  13. 13.
    Jahirabadkar, S., Kulkarni, P.: SCAF-An efficient approach to classify subspace clustering. Int. J. Data Mining Knowl. Manage. Process, 3(2) (2013)Google Scholar
  14. 14.
    Hand, D.J., Mannila, H., Smyth, P.: Principles of Data Mining, MIT Press, Cambridge, pp. 302–305 (2001)Google Scholar
  15. 15.
    Tan, P., Steinbach, M.K.: An Introduction to Data Mining. Wesley, London (2005)Google Scholar
  16. 16.
    A. Alfakih, A. Khandani, and H. Wolkowicz.: Solving Euclidean distance matrix completion problems via semide¯nite programming. Comput. Optim. Appl. 12, 13–30 (1999)Google Scholar

Copyright information

© Springer India 2015

Authors and Affiliations

  1. 1.AMET UniversityChennaiIndia
  2. 2.Department of Information TechnologyThakur College of Science and CommerceMumbaiIndia

Personalised recommendations