Scala and Spark for Big Data Analytics

Hierarchical clustering

The hierarchical clustering technique is based on the fundamental idea of objects or features that are more related to those nearby than others far away. Bisecting K-means is an example of such hierarchical clustering algorithm that connects data objects to form clusters based on their corresponding distance.

In the hierarchical clustering technique, a cluster can be described trivially by the maximum distance needed to connect parts of the cluster. As a result, different clusters will be formed at different distances. Graphically, these clusters can be represented using a dendrogram. Interestingly, the common name hierarchical clustering evolves from the concept of the dendrogram.