Broad intuition · Study map 3 of 85
Learning from labels, patterns, and feedback
Suggested preparation: Search, logic, and planning
Picture it
A label is an answer key. Unsupervised learning looks for structure without one. Self-supervised learning hides part of the data and asks the model to recover it; the data supplies its own answer key. Reinforcement learning uses consequences of actions.
What to understand
- Regression versus classification
- Supervised, unsupervised, self-supervised, and reinforcement learning
Follow the references
1 · Course
Google · Machine Learning Crash Course ↗Follow regression and classification, then contrast their targets with next-token prediction.
You understand it when…
For house prices, document clusters, missing words, and a game-playing agent, identify the training signal.
This is a study map. Follow the reference for the full explanation; use the question to test your mental model. Browse the reference shelf.