Machine Learning · Chapter 18 of 40
Hierarchical Clustering
Builds a tree (dendrogram) by successively merging the closest clusters.
No need to specify K upfront — cut the dendrogram at any level to get any number of clusters.
Example 1 (python)
from sklearn.cluster import AgglomerativeClustering
h = AgglomerativeClustering(n_clusters=3).fit(X)
print(h.labels_[:10])Output
[0 1 0 2 1 0 2 1 0 2]Cluster labels.
Example 2 (python)
# Plot dendrogram with scipy.cluster.hierarchy.dendrogramVisualise the merge tree.
Key points
- Builds a merge tree.
- No K needed to build.
- Slower than K-means.
- Dendrogram is very informative.
💡 Note: Agglomerative (bottom-up) is common; divisive (top-down) is rare in practice.