🌿 03 — Hierarchical Clustering

Hierarchical clustering builds a tree of clusters.

---

Agglomerative Clustering

Starts with each point as its own cluster, then merges similar clusters.

from sklearn.cluster import AgglomerativeClustering
from sklearn.preprocessing import StandardScaler

X_scaled = StandardScaler().fit_transform(X)

model = AgglomerativeClustering(n_clusters=3)
labels = model.fit_predict(X_scaled)

---

Dendrogram

SciPy can visualize merge structure:

from scipy.cluster.hierarchy import linkage, dendrogram
import matplotlib.pyplot as plt

linked = linkage(X_scaled, method="ward")
dendrogram(linked)
plt.show()

---

When Useful

• small to medium datasets
• when cluster hierarchy matters
• when you want to inspect possible cluster counts

---

Next

➡️ 04-dbscan