AI Rookies

SARSA — State-Action-Reward-State-Action

Fact

A reinforcement learning algorithm that updates value using the action it actually takes.

In Plain Words

SARSA is a driving instructor with a dashcam. If you turn into the taco drive-through, it grades that trip.

Robots and game bots use it. It learns by trying the current plan in real runs.

Related Concepts

RL
SARSA is a classic RL method for learning an action plan.

TD Learning
SARSA uses TD Learning to update value from the next step.

On-Policy Learning
SARSA learns from the action its current policy really takes.

Q-Learning
SARSA learns the real route, and Q-Learning learns the best-looking route.