High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

Clustering Multidimensional Data using High Density Point K-means Algorithm


G. Madhumitha , Angel college of Engineering and Technology; K. Kathiresan, Angel college of Engineering and Technology


Density Based Clustering, Improved K-Means, Improved Initial Centroids, Numeric Data Clustering, Similarity Index


Clustering is one of the major data analysis technique. It divides the data objects into several groups called clusters. K-means is one of the widely used partition based clustering technique. But k-means suffer from some limitations such as initial centroid selection and determining the number of clusters. The final clustering results heavily depends on initial centroids so proper selection of initial centroid is necessary. This paper introduces a new technique called High Density Point k-Means algorithm to start the k-means with better initial centroids. In this algorithm, the data objects with high density values is chosen as the initial cluster center. Experimental results shows that the proposed algorithm produces more accurate and efficient clusters. Finally, Cluster efficiency is measured by using similarity index.

Other Details

Paper ID: IJSRDV6I120411
Published in: Volume : 6, Issue : 12
Publication Date: 01/03/2019
Page(s): 619-621

Article Preview

Download Article