Chapter 09 · Sustainability & AISustainability & AI
Unsupervised learning
Definition
Machine learning that finds structure in unlabelled data — clustering similar items, reducing dimensions, or detecting anomalies — without being told the right answers.
References
This source provides part of the technical or institutional basis for the definition.
This source supports the explanation of how the term is applied, measured or governed in practice.
Overview
What it means
Clustering and dimensionality-reduction techniques group sites, suppliers, consumers or time series by similarity, revealing patterns nobody pre-defined. Results need human interpretation: the structure found is real, but what it means is an analytical judgement.
How it is used
Uses include segmenting energy consumers for demand programmes, grouping suppliers by risk profile, detecting unusual emissions readings, and exploring large environmental datasets before hypothesis formation.
Why it matters
Unsupervised methods are discovery tools for the data-rich, label-poor situations common in sustainability. Their freedom is also their hazard: patterns can be spurious, and clusters can quietly encode proxies for protected or sensitive characteristics.