• rashmi
• rainu nandal

Abstract: - K-means is the learning algorithm which is one of simplest and unsupervised. In this clustering algorithm k centers are defined for each cluster. These centers are placed such that each center is away from each other as far as possible. Now, associate each point which belongs to data set to the nearest center. This process of assigning data set is continued until no point is left behind. In the next step k new centroids are calculated again and a new binding is done again between the same data points. This step is repeated until there is no change in the position of data points. EM algorithm is expectation-maximization algorithm which is a method of finding maximum likelihood of parameters in models, which depends on latent variables [1]. This algorithm estimates the current parameters and using these it creates a function for the expectation to complete expectation step. The next step is the maximization step, in which those parameters are computed which is having maximum log-likelihood [1]. In this paper we will use the tool WEKA for performing clustering on the data points. In weka we will apply two different clustering algorithms on the real data set and try to analyse the two on the basis of no. clusters formed by them. At the end we will conclude the two as observed by the result of execution few data sets.

WEKA, K-means, EM, clustering.

"COMPARITIVE ANALYSIS OF CLUSTERING ALGORITHMS", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 3, page no.482-492, March-2019, Available :http://www.jetir.org/papers/JETIR1903A73.pdf

"COMPARITIVE ANALYSIS OF CLUSTERING ALGORITHMS", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 3, page no. pp482-492, March-2019, Available at : http://www.jetir.org/papers/JETIR1903A73.pdf