Definition
Unsupervised Learning is a Machine Learning method for finding patterns and structures in data without known target values or labels.
Human Explanation
The model searches for similarities, groups or unusual structures without examples of a predefined correct answer.
Why it Matters
It enables exploration of unlabelled datasets and can reveal structures that were not specified in advance.
Conceptual Boundary
Unsupervised Learning does not learn to reproduce a known target in the same way as Supervised Learning. Discovered groups or patterns are mathematical results that still require interpretation and validation.
Practical Perspective
Common uses include clustering, anomaly detection, dimensionality reduction and data exploration. Results can change substantially with data representation, scaling, method and parameters.