Agglomerative Clustering and Dendrogram Visualization using Scikit-Learn

 

Experiment

Agglomerative Clustering and Dendrogram Visualization using Scikit-Learn

---

🎯 Objective

To implement Agglomerative Hierarchical Clustering using scikit-learn and visualize the clustering using a dendrogram.

---

📘 Theory

🔹 Hierarchical Clustering

Hierarchical clustering creates a tree-like structure of clusters:

• Agglomerative (Bottom-Up) → merges closest clusters step-by-step

---

🔹 Agglomerative Clustering in Scikit-Learn

The class AgglomerativeClustering:

• Does not require manual distance computation
• Supports linkage methods:
  • single
  • complete
  • average
  • ward (default)

---

🔹 Linkage Methods

Method	Description
Single	Minimum distance
Complete	Maximum distance
Average	Mean distance
Ward	Minimizes variance

---

🔹 Dendrogram

• Visual representation of cluster merging
• Height = distance between clusters
• Helps decide number of clusters

---

🧾 Sample Dataset

Point	X	Y
A	1	1
B	2	1
C	4	3
D	5	4
E	8	7

---

💻 Program (Python Code)

import numpy as np
import matplotlib.pyplot as plt
from sklearn.cluster import AgglomerativeClustering
from scipy.cluster.hierarchy import dendrogram, linkage

# -------------------------------
# Step 1: Dataset
# -------------------------------
X = np.array([
    [1, 1],
    [2, 1],
    [4, 3],
    [5, 4],
    [8, 7]
])

labels = ['A', 'B', 'C', 'D', 'E']

# -------------------------------
# Step 2: Apply Agglomerative Clustering
# -------------------------------
model = AgglomerativeClustering(
    n_clusters=3,        # choose number of clusters
    linkage='single'     # same as manual experiment
)

cluster_labels = model.fit_predict(X)

# -------------------------------
# Step 3: Plot Clusters
# -------------------------------
plt.figure()
plt.scatter(X[:, 0], X[:, 1], c=cluster_labels)

for i, txt in enumerate(labels):
    plt.text(X[i][0]+0.1, X[i][1]+0.1, txt)

plt.title("Agglomerative Clustering (Scikit-Learn)")
plt.xlabel("X")
plt.ylabel("Y")
plt.show()

# -------------------------------
# Step 4: Dendrogram
# -------------------------------
Z = linkage(X, method='single')

plt.figure()
dendrogram(Z, labels=labels)

plt.title("Dendrogram (Scipy)")
plt.xlabel("Data Points")
plt.ylabel("Distance")
plt.show()

# -------------------------------
# Step 5: Output
# -------------------------------
print("Cluster Labels:", cluster_labels)

---

📊 Output

🔹 Cluster Labels (Example)

Cluster Labels: [2 2 0 0 1]

---

🔹 Cluster Interpretation

• Cluster 1 → A, B, C, D
• Cluster 2 → C,D
• Cluster 3 → E

---

🔹 Dendrogram

• Shows merging sequence:
1. A + B
2. C + D
3. (AB) + (CD)
4. (ABCD) + E

---

📌 Result

Agglomerative clustering was successfully performed using scikit-learn, and the hierarchical structure was visualized using a dendrogram.

• Library implementation simplifies clustering
• Dendrogram provides clear visualization
• Results match manual implementation