Agglomerative Clustering
Explore the fundamentals of agglomerative clustering, a bottom-up hierarchical technique that merges data points into clusters based on similarity. Understand key linkage methods such as single, complete, and centroid distance, and learn how the algorithm iteratively combines clusters to form a dendrogram representing nested groupings.
Hierarchical clustering is a machine learning technique that creates nested clusters by iteratively merging or dividing data points. In this lesson, we focus specifically on agglomerative clustering, which is a bottom-up hierarchical approach. We will walk through the core process of agglomerative clustering, explain the main methods used to define similarity or dissimilarity between clusters (known as linkage), and outline the steps of the agglomerative clustering algorithm.
The agglomerative process
Agglomerative clustering starts by treating each data point as its own cluster. At every step, the algorithm finds the two closest clusters and merges them. This process continues until only the desired number of clusters ...