Which clustering method creates a multi-level hierarchy of clusters?

Which clustering method creates a multi-level hierarchy of clusters?

• A. Hierarchical Clustering
• B. Agglomerative Clustering
• C. Divisive Clustering
• D. Density-based Clustering

Answer: (A)

Creating a multi-level hierarchy of clusters is the hallmark of hierarchical clustering. It builds a tree-like structure (a dendrogram) that shows how small groups nest inside larger ones, revealing structure at different levels. You can view different granularities by cutting the tree at various heights, which is exactly what a multi-level hierarchy provides. There are two common ways to build this hierarchy: agglomerative, which starts with each data point as its own cluster and successively merges them, and divisive, which starts with all points in one cluster and splits them step by step. Both are forms of hierarchical clustering, so they inherently produce nested cluster structures. Density-based clustering, such as DBSCAN, groups points by local density and tends to form flat clusters rather than a nested hierarchy. It doesn’t inherently yield multi-level hierarchical clusters, which is why hierarchical clustering is the correct choice here.

Creating a multi-level hierarchy of clusters is the hallmark of hierarchical clustering. It builds a tree-like structure (a dendrogram) that shows how small groups nest inside larger ones, revealing structure at different levels. You can view different granularities by cutting the tree at various heights, which is exactly what a multi-level hierarchy provides.

There are two common ways to build this hierarchy: agglomerative, which starts with each data point as its own cluster and successively merges them, and divisive, which starts with all points in one cluster and splits them step by step. Both are forms of hierarchical clustering, so they inherently produce nested cluster structures.

Density-based clustering, such as DBSCAN, groups points by local density and tends to form flat clusters rather than a nested hierarchy. It doesn’t inherently yield multi-level hierarchical clusters, which is why hierarchical clustering is the correct choice here.