A Detailed Study and Analysis of different Partitional Data Clustering Techniques

The concept of Data Clustering is considered to be very significant in various application areas like text mining, fraud detection, health care, image processing, bioinformatics etc. Due to its application in a variety of domains, various techniques are presented by many research domains in the literature. Data Clustering is one of the important tasks that make up Data Mining. Clustering can be classified into different types such as partitional, hierarchical, spectral, density-based, grid-based, model based etc. Among the different types of clustering available, partitional clustering is the most widely used in most of the applications since the computation involved is not very complex. Hence lot of research has been carried out in clustering using partitional method. In this paper, it is proposed to do a comprehensive study of the different partitional clustering techniques used in the literature which will also provide an insight into the recent problems in the same area. In this paper, sixteen research articles have been taken which are published by different publishers between the years 2005 and 2013. Various algorithms come under partitional clustering among which Bisecting K-Means is an excellent one that gives a good quality output for clustering large number of data. Also a broad analysis is carried out to provide an insight into the importance of the various approaches which can in turn throw light to developments in the same area.