← Back to resources

Unsupervised Learning

Unsupervised learning discovers hidden structure in data through methods such as clustering and representation learning.

Definition

Unsupervised learning is a type of machine learning in which models identify patterns in data without labelled training examples.

How It Works in AI

Instead of learning from predefined answers, unsupervised methods discover hidden structure in data using techniques such as clustering, dimensionality reduction, and representation learning. This lets models group similar items and build internal representations of meaning from raw data.

Unsupervised learning is central to modern AI systems and language model training because massive unlabelled corpora are easier to obtain than fully annotated datasets. It provides the foundation for strong embeddings and broad generalisation.

It is foundational in modern AI systems and language model training, where large-scale unlabelled data drives representation quality.

Related Terms

Universal Language Models

Multilingual language models that use shared representations to support many languages and NLP tasks.

Accountability in AI

Organisational responsibility for how AI systems function, make decisions, and impact users.

Algorithmic Bias

Systematic errors in AI outputs arising from skewed, imbalanced, or prejudiced training data.

Related Resources

Supervised Learning

A machine learning method in which models are trained using labelled data.

Federated Learning

A machine learning approach where models are trained across distributed devices or servers without centralising raw data.

Machine Learning

A field of AI where systems learn patterns from data to make predictions or generate content.

Overfitting

A machine learning problem where a model learns the training data too closely and performs poorly on new or unseen inputs.

Deep Learning

A machine-learning approach using multi-layer neural networks to model complex patterns.

Gender Bias in AI

Differences in how AI treats or represents genders due to training-data patterns.

Explore Trad AI

Open the workspace