Tutorial 28: Hierarchical Clustering

Level: Intermediate · Part of: ML Learning Roadmap

← Back to roadmap

Case Study: Country Development Grouping

Scenario

The UN wants to group countries by development indicators. Hierarchical clustering builds a dendrogram showing nested group structure.

Learning Objectives

• Use AgglomerativeClustering
• Visualize dendrograms with scipy
• Compare linkage methods (ward, complete, average)
• Cut dendrogram at desired k

Prerequisites

• Tutorial 27
• Python 3.9+, scikit-learn, NumPy, pandas, matplotlib

Dataset

load_iris() for demonstration

Hands-On Solution

Copy and run the complete script below:

import matplotlib.pyplot as plt
import numpy as np
from scipy.cluster.hierarchy import dendrogram, linkage
from sklearn.cluster import AgglomerativeClustering
from sklearn.datasets import load_iris
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import adjusted_rand_score

iris = load_iris()
X = StandardScaler().fit_transform(iris.data)

# Dendrogram (subsample for readability)
idx = np.random.RandomState(42).choice(len(X), 50, replace=False)
Z = linkage(X[idx], method="ward")

plt.figure(figsize=(10, 5))
dendrogram(Z, truncate_mode="level", p=4)
plt.title("Hierarchical Clustering Dendrogram (Ward)")
plt.savefig("dendrogram.png", dpi=120)

for method in ["ward", "complete", "average"]:
    agg = AgglomerativeClustering(n_clusters=3, linkage=method)
    labels = agg.fit_predict(X)
    ari = adjusted_rand_score(iris.target, labels)
    print(f"linkage={method:10s}  ARI={ari:.3f}")

Expected Output

When you run the script, you should see evaluation metrics printed to the console. Some tutorials also save .png plot files in the current directory.

Exercises

1. Cut the dendrogram at 4 clusters. How does ARI change?
2. Compare hierarchical vs K-Means on iris (use ARI).
3. When is hierarchical clustering preferred over K-Means?

Key Takeaways

• Hierarchical clustering needs no preset k (but you still choose a cut)
• Ward linkage minimizes within-cluster variance
• Dendrograms provide rich visual structure but scale poorly to large n

Navigation

Part of the ML Learning Roadmap — Hands-On with scikit-learn