Clustering is an unsupervised machine learning technique that automatically groups similar data points together, without being told in advance what the groups should be.
What it means in plain English
Clustering finds natural groupings in data. You give it a set of items and it sorts them into clusters, where items in the same cluster are more similar to each other than to those in other clusters. Because it needs no labels, it is ideal for discovering structure you did not know was there — revealing categories the data forms on its own.
It is one of the most common unsupervised learning tasks.
A simple example
A business with no predefined customer types can use clustering to automatically group its customers by purchasing behaviour, uncovering natural segments — such as frequent small buyers versus occasional big spenders — to target differently.
Why it matters
Clustering powers customer segmentation, organising large collections of documents or images, and exploratory data analysis. Its ability to find hidden groups makes it a key tool for making sense of unlabelled data.
Related terms
- Unsupervised Learning — the category clustering belongs to.
- Classification — the supervised counterpart with known categories.
- Embedding — often clustered to find similar items.
Frequently asked questions
What is clustering used for?
Grouping similar items without predefined labels — for example, segmenting customers by behaviour, grouping documents by topic, or finding patterns in data during exploration.
Is clustering supervised or unsupervised?
Clustering is unsupervised: it finds structure in unlabeled data by grouping similar points together, rather than learning from labelled examples.
