ABSTRACT
The dendrogram is a mathematical, as well as a visual representation of a hierarchical procedure that can be divisive or agglomerative. The stem or edge shows children of internal nodes and the connection with the clusters below it. The length of the edge represents the distances at which clusters are joined. The dendrograms for hierarchical clustering are binary trees; so, they have two edges emanating from each internal node. The topology of the tree refers to the arrangement of stems and nodes. Dendrograms are very familiar to data analysts working in hierarchical clustering applications, and they are easy to understand because they match the concept of how trees are laid out in a physical sense with branches and leaves. The ReClus method was developed by Martinez as a way to view the output of nonhierarchical clustering methods, such as k-means, model-based clustering, etc., that is reminiscent of the treemap and rectangle displays.
