Reinforcement learning
Reinforcement learning studies agents that improve decisions through interaction and reward. The objective is expected return over time, not simply predicting a labeled answer.
Definition
Reinforcement learning studies agents that improve decisions through interaction and reward. The objective is expected return over time, not simply predicting a labeled answer.
Intuition
An action can change both the immediate payoff and the situations the agent sees next. That delayed effect distinguishes a sequential problem from a one-step prediction.
Worked example
A robot may take a longer route now to reach a charging station before attempting a delivery.
The math
A policy maps a state to an action distribution: . An agent seeks to maximize expected cumulative reward.
In practice
Use RL when actions affect future opportunities and you can evaluate a reward safely. Start with a simple baseline.
Go deeper
Sources
More in Reinforcement learning
Assembled from the ReLU.chat curated knowledge base. These explanations are concise on purpose; check the sources for anything important.