• Kinjal Thakar

The clustering techniques are the most important part of the data analysis and k-means is the oldest and popular clustering technique used. The paper discusses the traditional K-means algorithm with advantages and disadvantages of it. It also includes researched on enhanced k-means proposed by various authors and it also includes the techniques to improve traditional K-means for better accuracy and efficiency. There are two area of concern for improving K-means; 1) is to select initial centroids and 2) by assigning data points to nearest cluster by using equations for calculating mean and distance between two data points. The time complexity of the proposed K-means technique will be lesser that then the traditional one with increase in accuracy and efficiency. The main purpose of the article is to proposed techniques to enhance the techniques for deriving initial centroids and the assigning of the data points to its nearest clusters. The clustering technique proposed in this paper is enhancing the accuracy and time complexity but it still needs some further improvements and in future it is also viable to include efficient techniques for selecting value for initial clusters(k). Experimental results show that the improved method can effectively improve the speed of clustering and accuracy, reducing the computational complexity of the k-means.

Data Analysis, Clustering, k-means Algorithm, Clustering Algorithm

"IMPROVED K-MEANS ALGORITHM BASED ON DISSIMILAR VALUES", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.5, Issue 7, page no.1658-1661, July-2018, Available :http://www.jetir.org/papers/JETIRC006285.pdf

"IMPROVED K-MEANS ALGORITHM BASED ON DISSIMILAR VALUES", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.5, Issue 7, page no. pp1658-1661, July-2018, Available at : http://www.jetir.org/papers/JETIRC006285.pdf