Hierarchical clustering groups data items based on different levels of a hierarchy. It combines the items in groups based on different levels of a hierarchy using top-down or bottom-up strategies. Based on the strategy used, hierarchical clustering can be of two types – agglomerative or divisive:
- The agglomerative type is the most widely used hierarchical clustering technique. It groups similar data items in the form of a hierarchy based on similarity. This method is also called Agglomerative Nesting (AGNES). This algorithm starts by considering every data item as an individual cluster and combines clusters based on similarity. It iteratively collects small clusters and combines them into a single large cluster. This algorithm gives its result in the form of a tree structure. It works in a bottom-up manner; that is, every item is initially considered as a single element cluster and in each iteration of the algorithm, the two most similar clusters are combined...
