IJIRST (International Journal for Innovative Research in Science & Technology)ISSN (online) : 2349-6010

 International Journal for Innovative Research in Science & Technology

Accelerating Unique Strategy for Centroid Priming in K-Means Clustering


Print Email Cite
International Journal for Innovative Research in Science & Technology
Volume 3 Issue - 7
Year of Publication : 2016
Authors : Ms. S. Saranya ; Ms. P. Deepika; Ms. S. Sasikala; Dr. S. Jansi; Ms. A. Kiruthika

BibTeX:

@article{IJIRSTV3I7013,
     title={Accelerating Unique Strategy for Centroid Priming in K-Means Clustering},
     author={Ms. S. Saranya, Ms. P. Deepika, Ms. S. Sasikala, Dr. S. Jansi and Ms. A. Kiruthika},
     journal={International Journal for Innovative Research in Science & Technology},
     volume={3},
     number={7},
     pages={40--47},
     year={},
     url={http://www.ijirst.org/articles/IJIRSTV3I7013.pdf},
     publisher={IJIRST (International Journal for Innovative Research in Science & Technology)},
}



Abstract:

Clustering is the process of organizing data objects into a set of disjoint classes called clusters. The fundamental data clustering problem may be defined as discovering groups in data or grouping similar objects together. Some of the problems associated with current clustering algorithms are that they do not address all the requirements adequately, and need large number of iterations when dealing with a large number of dimensions. K-Means is one of the algorithms that solve the well-known clustering problem. This algorithm classifies object to a predefined number of clusters, which is given by the user. The idea is to choose random cluster centers, one for each other. The centroid initialization plays an important role in determining the cluster assignment in effective ways. But the performance of K-Means clustering is affected when the dataset used is of high dimension and the accuracy and sum square error is highly affected because of the high dimension data. This paper, proposed a new algorithm of data partitioning based k-means for performing data partitioning along the data axis with the highest variance. This will shows more effective and efficient converge to better clustering results, reduce the number of iterations required clustering also help to reduce the sum square error for all cells than the existing clustering.


Keywords:

Data clustering, k-means algorithm, Data partitioning


Download Article