Off-policy-learning — Off-Policy Learning

Fact

A way for AI to learn from data made by another policy.

In Plain Words

Off-policy learning is like studying Mario Kart from your cousin’s messy replays. You did not drive those laps, but you learn when to brake.

The AI learns from old runs or other players’ moves. It shows up in recommenders, games, and robots.

Related Concepts

• Q-Learning
• Reinforcement Learning
• Deep Q-Network
• Temporal-Difference Learning

Q-Learning
Q-Learning is the classic example of off-policy learning.

RL
Off-policy learning is part of RL and can use old experience.

Deep Q-Network
Deep Q-Network uses replay data to learn from old runs again.

TD Learning
Off-policy learning often uses TD Learning to update value guesses.