Data Science and Machine Learning Interview Handbook/

...

/

Hierarchical Clustering

Hierarchical Clustering

Explore implementation for a hierarchical clustering algorithm.

We'll cover the following...

Hierarchical clustering is a powerful technique for discovering nested structures in data, often revealing hidden patterns that flat clustering methods can miss. In this lesson, we’ll build a hierarchical clustering workflow, visualize the results using dendrograms, and compare different distance metrics for clustering quality. Let’s get started.

Hierarchical clustering implementation

Hierarchical clustering is a popular unsupervised learning algorithm we use within our company. It helps us identify natural groupings within data, which can be crucial for uncovering hidden patterns and insights.

Implement a simple hierarchical clustering algorithm that performs linkage and creates a diagram, given sample data. Your implementation should be efficient, can leverage scipy, and needs to visualize the dendrogram for a sample dataset.

Press + to interact
Python 3.10.4
import numpy as np
from scipy.cluster.hierarchy import dendrogram, linkage
from sklearn.datasets import make_blobs
def perform_hierarchical_clustering(X):
    # TODO: Implement hierarchical clustering
    # 1. Perform linkage 
    # 2. Create dendrogram
    # Hint: Use linkage() and dendrogram() from scipy.cluster.hierarchy
    
    # Your implementation here
    pass
# Generate sample data
X, _ = make_blobs(n_samples=50, centers=3, random_state=42)
# Call your function
perform_hierarchical_clustering(X)

Sample answer

Here’s how we can break this down:

  1. Prepare the data: Normalize features if they are on different scales using StandardScaler or MinMaxScaler.

  2. Perform linkage: Use ...