Hierarchical cluster analysis is an unsupervised machine learning and statistical method that organizes similar data points into a nested hierarchy of clusters. The approach constructs this tree-like structure through either an agglomerative strategy, where individual data points start in their own clusters and iteratively merge based on proximity, or a divisive strategy, where all data points begin in a single cluster that is recursively divided. Pairwise distances and linkage criteria govern how these clusters combine or split. Unlike partitioning methods that require the target number of groups to be specified in advance, hierarchical cluster analysis generates a multi-level structure often visualized as a dendrogram, allowing analysts to explore data relationships at varying degrees of granularity.
