Hierarchical clustering techniques approach the analysis a bit differently than k-means clustering. Instead of working with a predetermined number of centers and iterating to find membership, hierarchical techniques continually pair or split data into clusters based on similarity (distance). There are two different approaches:
Divisive clustering: This begins with all the data in a single cluster and then splits it and all subsequent clusters until each data point is its own individual cluster
Agglomerative clustering: This begins with each individual data point and pairs them together in a hierarchy until there is just one cluster
In this section, you will learn and use agglomerative hierarchical clustering. It is a bit faster than divisive clustering, but they both may work slow with very large datasets. One benefit of hierarchical approaches is that they do not require you to specify the number of clusters in advance. You can run the model and prune...