Abstract

K-means algorithm is one of the most famous clustering algorithms in data mining due to its simplicity. Kernel K-means is an extension of K-means to cluster nonlinear separable data. However, it still has some limitations like sensitivity and convergence to the local optima. In this paper, we show how to implement a new novel kernel-clustering algorithm that is robust and converges to the global solution. We show using artificial and real data sets that the proposed kernel algorithm performs better than the standard kernel K-means algorithm.

References

• Jain, A.K., M.N. Murty and P.J. Flynn, 1999. Data clustering: A review. ACM Comput. Surv., 31: 264-323.
• Xindong, Wu and et. al 2008. Top 10 Algorithms in Data Mining. Journal of Knowledge and Information Systems, 14(1):1-37, DOI: 10.1007/s10115-007-0114-2.
• Plant, C. ; Zherdin, A. ; Sorg, C. ; Meyer-Baese, A. ; Wohlschlager, A.M., 2014, Mining Interaction Patterns among Brain Regions by Clustering, IEEE Transactions on Knowledge and Data Engineering, 26(9): 2237-2249, DOI: 10.1109/TKDE.2013.61.
• Shuo Chen and Chengjun Liu, 2014, Clustering-Based Discriminant Analysis for Eye Detection, IEEE Transactions on Image Processing, 23(4):1629-1638, DOI: 10.1109/TIP.2013.2294548.
• Guojun Gan, Chaoqun Ma, and Jianhong Wu, 2007. Data Clustering: Theory, Algorithms, and Applications, ISBN: 978-0-898716-23-8, ASA-SIAM.
• Celebi, M., H Kingravi, and P. A. Vela, 2013, “A comparative study of efficient initialization methods for the k-means clustering algorithm.” Expert Syst. Appl., vol. 40:200–210.
• Cao, F. Y., Liang, J. Y. and Jiang, G., 2009. An initialization method for the K-means algorithm using neighborhood model. Computers and Mathematics with applications, 58(3): 474-483.
• Likas, A., Vlassis,, M. and Verbeek, J., 2003, “The global k-means clustering algorithm,” Pattern Recognition, vol. 36, pp. 451–461.
• Arai, Koheri and Ridho, Ali, 2007. Hierarchical K-means, an algorithm for Centroids initialization for K-means, Saga University, 36(1): 25-31.
• Ashour W., Fyfe, C., 2008, Local vs global interactions in clustering algorithms: advances over K-means, International Journal of Knowledge-based and Intelligent Engineering Systems (KES), 12(2): 83-99, 2008. ISSN 1327-2314.
• Khan, S., Ahmad, A., 2004, Cluster center initialization algorithm for K-means clustering. Pattern Recognition Letters, vol. 25, pp.1293-1302.
• Arthur, D., and Vassilvitskii, S., 2006. K-means++: The advantages of careful seeding. In Bay Area Theory Symposium,BATS06. http://www.stanford.edu/~sergeiv/papers/kMeansPP-soda.pdf.
• Zhang, B., Hsu, M., and Dayal, U., 1999. K-harmonic means - a data clustering algorithm. Technical Report HPL-1999-124, HP Laboratories, Palo Alto.
• B. Zhang, 2001. Generalised k-harmonic means- dynamic weighting of data in unsupervised learning. In First SIAM International Conference on Data Mining. http://www.siam.org/meetings/sdm01/pdf/sdm01_06.pdf.
• Angelov, P., 2004. An approach for fuzzy rule-base adaptation using on-line clustering. International Journal of Approximate Reasoning, 35(3):275–289.
• Dhillon, S., Guan, Y., and Kulis, B, 2004. Kernel k-means, spectral clustering and normalized cuts. In Proc. ACM SIGKDD Intl Conf. Knowledge Discovery and Data Mining, Seattle, W.
• Girolami, M.., 2002, Mercer kernel based clustering in feature space IEEE Transactions on Neural Networks, 13(3):780- 784.
• Burges C. 1998. A tutorial on support vector machines for pattern recognition. Data Mining and Knowledge Discovery, 2(2):121–167.
• Suykens, J. A. and J. Vandewalle, J. 1999. Least squares support vector machine classifiers. Neural Processing Letters, 9(3):293–300.
• Ashour, W., Wu, Y. and Fyfe, C, 2009. Non-standard parameter adaptation for exploratory data analysis, Springer.