Glossary

Clustering

An unsupervised technique that groups data points so items within a group are more similar to each other than to items outside it.

Also called: cluster analysis

Clustering is an unsupervised learning technique that groups data points into clusters based on similarity, without using any pre-labeled outcome. It is used to discover structure in data rather than to predict a known target.

Algorithms differ in how they define "similar." k-means clustering assigns points to a fixed number of clusters by minimizing distance to each cluster's center; hierarchical clustering builds a tree of nested groupings; density-based methods find clusters of arbitrary shape and flag sparse points as outliers. Because there is no ground truth to check against, clustering quality is judged with internal metrics, such as how tight and well-separated the clusters are, or with domain judgment.

In analytics, clustering is the basis for customer segmentation, grouping similar users, transactions or documents, and surfacing structure that would be hard to see manually. It is often preceded by dimensionality reduction, since clustering on many raw, correlated features tends to produce noisy groups. A common pitfall is treating the resulting clusters as objectively "real" categories rather than one of many valid ways to partition the same data, and choosing the number of clusters arbitrarily rather than checking whether it is actually supported by the data.

Last reviewed September 22, 2026

In the index now

Related terms

Unsupervised learning A machine learning approach that finds structure or patterns in data that has no labeled correct answer to learn from. K-means clustering A clustering algorithm that partitions data into k groups by repeatedly assigning points to the nearest of k cluster centers. Dimensionality reduction Techniques that compress a dataset with many variables into fewer dimensions while preserving as much meaningful information as possible. Customer segmentation Dividing a customer base into groups that share meaningful traits so each can be targeted or served differently. Anomaly detection Identifying data points or patterns that deviate significantly from expected or normal behavior.

Related guides

How to segment customers RFM, behavioral and clustering-based segmentation compared, and what turns a segment from an analysis artifact into something teams actually use.