Hierarchichal clustering

Imagine sorting a basket of fruits. Clustering groups similar fruits based on features like color, size, and shape.

What is Clustering?

Hierarchical clustering starts with individual data points and progressively merges them into clusters based on similarity.

The Hierarchy

Agglomerative: Starts with single points. Divisive: Starts with all data in one cluster.

Two Main Approaches

To measure similarity, we use distance metrics like Euclidean distance (straight line) or Manhattan distance (taxicab geometry).

Choosing the Right Distance

A dendrogram, a tree-like structure, shows the merging process and the relationships between clusters at different levels.

Visualizing the Hierarchy

– Customer segmentation – Image analysis – Document clustering

Real-World Applications