Clustering

Clustering is dividing data into groups such that every group has similar data. It is basically a collection of data on the basis of similarities and dissimilarities between them. Clustering is important since it regulates the essential grouping among the unlabelled data. Where there is no measure for good clustering, it depends on in what way the users use the data. Density-​based method, hierarchical-​based methods, portioning methods and grid-​based methods are the clustering methods. Density-​based methods consider the cluster as the compressed region having some similarity as differentiated from the compressed region of the rest of the space. These methods have high accuracy and capacity to unite two clusters. Hierarchical-​based methods form a tree-​type structure based on hierarchy. New clusters are formed using earlier ones. These methods are divided into two categories: agglomerative and divisive.