Back to Glossary
BasicsAI Glossary

Deep Learning

Quick Answer

Deep Learning is a branch of machine learning that uses multi-layered neural networks to learn highly complex patterns directly from raw data such as images, audio and text, without the need for hand-crafted feature engineering.

In Depth

What Deep Learning really means

Deep Learning models contain many stacked layers of artificial neurons; each layer transforms its inputs into a more abstract representation. Deeper architectures can capture intricate relationships — such as recognising faces or translating languages — that shallower models cannot.

Modern deep learning underpins computer vision systems, speech recognition, autonomous vehicles and large language models. Training these models typically requires GPUs or specialised accelerators and very large datasets, making compute cost a key consideration for organisations.

Why It Matters

Business relevance for UK organisations

Deep learning is the right tool when you have unstructured data at scale — photographs, audio recordings, documents, video — and the problem is genuinely non-linear. It is overkill for small tabular datasets, where classical ML usually wins.

Real-world example

How this shows up in practice

A Leeds-based medical imaging startup uses deep learning to detect early signs of diabetic retinopathy from retina photographs, flagging high-risk patients for NHS referral.