Unsupervised Learning
Finding patterns, structure, or representations in data without labelled examples.
Definition
Unsupervised learning discovers hidden structure in unlabelled data. Without explicit supervision, models learn by finding natural groupings (clustering), reducing dimensionality (compression), or modelling the data distribution (generative modelling).
Key approaches include: clustering (k-means, DBSCAN, hierarchical) — grouping similar data points; dimensionality reduction (PCA, t-SNE, UMAP) — finding compact representations; density estimation — modelling the probability distribution of data; and autoencoders — learning compressed latent representations.
Modern self-supervised learning (a special form) has largely superseded classical unsupervised learning for representation learning, as it provides much richer supervisory signal. Classical unsupervised methods remain useful for exploratory data analysis, anomaly detection, and customer segmentation.
Examples
- Customer segmentation (k-means)
- Anomaly detection (autoencoders)
- Dimensionality reduction (t-SNE)
- Generative modelling (VAE)
Related Terms
Machine Learning (ML)
A subset of AI where systems learn patterns from data rather than following explicitly programmed rules.
Supervised Learning
Training ML models on labelled input-output pairs to learn a function mapping inputs to target outputs.
Self-Supervised Learning
Learning representations from unlabelled data by creating supervisory signals from the data itself.