Patterns and categories
Be able to sort objects into groups based on properties, explain the rule governing the group, and identify when something does not fit.
Everyday explanation
When you sort email messages into folders, you sort by a property: the subject, the sender, or the priority. You could have sorted by date instead. Same messages — different rule.
Finding patterns and dividing things into categories is the foundation of how computers "learn". The system looks at properties and groups information.
Intuition
A category needs a clear rule: "All invoices from supplier A" or "All messages with the word 'urgent'". With the same rule, the same objects always end up in the same group.
Sometimes something does not quite fit — an email might be from a supplier but concern a private matter. Then the rule is too simple. Improving the rule is the core of machine learning, except the computer does it with thousands of examples.
Mastery means
- Sorts objects according to a specified property
- Formulates the rule behind a given sorting
Sign in to do the exercises and build your mastery up.
Sources
Leads to
Part of the goals (42)
- Statistics for experiments
- Data: collect, clean, document
- Discoverer — build a game and train a machine
- Explorer — discover that computers can learn from examples
- Computers that see and hear
- From blocks to Python
- Train an agent with reward
- Build an AI service that survives production
- AI for beginners
- The mathematics behind the models
- Classical machine learning in practice
- Build an NLP system end to end
- The developer's toolbox
- AI, ethics and society
- Fine-tune and run your own models
- An AI service in operation
- Language models in practice
- Builder — collect data, train a model and test AI critically
- Training neural networks for real
- Classical ML for real
- Understand how generative AI works
- AI in production
- Frontier Lab — an independent research project
- Fine-tune a model with LoRA
- Build a RAG pipeline with traceable sources
- Generative models in depth
- Evals in practice
- Train your first neural network
- Responsible AI in practice
- Interpreting a language model
- AI safety in practice
- Deep reinforcement learning
- Build a voice interface
- Reproduce a paper
- Build a memory system for an agent
- Build an agent you can trust
- Seeing and hearing with AI
- The EU AI Act and GDPR in practice
- Multimodal systems
- Run models more cheaply: quantisation
- Image classification with convolutional networks
- Build a transformer from scratch