Hierarchical Clustering — Technology Wiki

Overview

Direct Answer

Hierarchical clustering is an unsupervised learning method that organises data points into a nested tree structure (dendrogram) by iteratively merging similar clusters (agglomerative approach) or splitting heterogeneous clusters (divisive approach). Unlike partitioning methods such as K-means, it does not require pre-specifying the number of clusters.

How It Works

Agglomerative hierarchical clustering begins with each data point as a singleton cluster, then sequentially merges the two closest clusters using a linkage criterion—such as single linkage (minimum distance), complete linkage (maximum distance), or average linkage (mean distance)—until a single encompassing cluster remains. The process generates a dendrogram that visualises cluster relationships at all granularities, allowing analysts to cut the tree at any height to obtain a desired number of clusters.

Why It Matters

Organisations value hierarchical clustering for exploratory data analysis because it reveals underlying cluster structure without prior assumptions, supports dendrogram-based decision-making, and scales naturally across domains from genomics to customer segmentation. The interpretability of the dendrogram aids stakeholders in validating cluster quality and understanding relationships between groups.

Common Applications

Applications include biological taxonomy and gene expression analysis in bioinformatics, document organisation and text mining in information retrieval, customer segmentation in retail and finance, and ecological species classification. Dendrograms are widely used in phylogenetic analysis and hierarchical taxonomy construction.

Key Considerations

Computational complexity grows quadratically with dataset size, making the method impractical for very large datasets. Linkage choice significantly influences results; greedy merging decisions are irreversible, potentially trapping the algorithm in suboptimal configurations.

Cross-References(1)

Machine Learning

Clustering

Related in Unsupervised Learning

Dimensionality Reduction

Techniques that reduce the number of input variables in a dataset while preserving essential information and structure.

Principal Component Analysis

A dimensionality reduction technique that transforms data into orthogonal components ordered by the amount of variance they explain.

t-SNE

t-Distributed Stochastic Neighbour Embedding — a technique for visualising high-dimensional data in two or three dimensions.

UMAP

Uniform Manifold Approximation and Projection — a dimensionality reduction technique for visualisation and general non-linear reduction.

Clustering

Unsupervised learning technique that groups similar data points together based on inherent patterns without predefined labels.

K-Means Clustering

A partitioning algorithm that divides data into k clusters by minimising the distance between points and their cluster centroids.

DBSCAN

Density-Based Spatial Clustering of Applications with Noise — a clustering algorithm that finds arbitrarily shaped clusters based on density.

Association Rule Learning

A method for discovering interesting relationships and patterns between variables in large datasets.

Collaborative Filtering

A recommendation technique that makes predictions based on the collective preferences and behaviour of many users.

Content-Based Filtering

A recommendation approach that suggests items similar to those a user has previously liked, based on item attributes.

Matrix Factorisation

A technique that decomposes a matrix into constituent matrices, widely used in recommendation systems and dimensionality reduction.

More in Machine Learning

Model Serving

MLOps & Production

The infrastructure and processes for deploying trained machine learning models to production environments for real-time predictions.

Ensemble Learning

MLOps & Production

Combining multiple machine learning models to produce better predictive performance than any single model.

Catastrophic Forgetting

Anomaly & Pattern Detection

The tendency of neural networks to completely lose previously learned knowledge when trained on new tasks, a fundamental challenge in continual and multi-task learning.

Class Imbalance

Feature Engineering & Selection

A situation where the distribution of classes in a dataset is significantly skewed, with some classes vastly outnumbering others.

Stochastic Gradient Descent

Training Techniques

A variant of gradient descent that updates parameters using a randomly selected subset of training data each iteration.

Batch Learning

MLOps & Production

Training a machine learning model on the entire dataset at once before deployment, as opposed to incremental updates.

Underfitting

Training Techniques

When a model is too simple to capture the underlying patterns in the data, resulting in poor performance on both training and test data.

Feature Engineering

Feature Engineering & Selection

The process of using domain knowledge to create, select, and transform input variables to improve model performance.