Clustering is a form of unsupervised learning that partitions observations into classes or clusters (collectively, called a clustering). An objective function or quality measure guides this search, ideally for a clustering that is optimal as measured by the objective function. A hierarchical-clustering system creates a tree-structured clustering, where sibling clusters partition the observations covered by their common parent. This section briefly summarizes a simple strategy, called hierarchical sorting, for creating hierarchical clusterings.
JAIR, 4