What does reinforcement learning mean?
Learning from rewards rather than labels: the program tries actions, is told how well each went, and shifts towards the actions that scored best.
In a lesson
No labels this time. Nobody tells the robot what the right action is. It tries things, and afterwards something tells it how well that went: a reward. Driving forward pays a little. Turning costs a little. Bumping into something costs a lot. From nothing but that, the robot works out how to wander a cluttered mat without hitting anything. This is reinforcement learning, and the method is called Q-learning.
Learn it
- Taught in 8.5 Learning by reward Learning, Robot club
Used in
- F13.1 How your exam works
- A14.1 The systems lifecycle and feasibility
- A15.7 Calculation and theory questions
- A15.8 A revision plan and mixed practice
- U12.1 Learning a behaviour
- U12.5 Reward is a specification
- U12.7 Capstone: the whole robot