Glossary term
Glossary term
Training and Fine-Tuning
Reinforcement learning is a method where AI learns by trial and error, getting rewarded for good outcomes and penalized for bad ones. It is useful for training agents to improve over time in dynamic or goal-driven environments.
DeepMind AlphaGo defeated Lee Sedol in 2016 using reinforcement learning with self-play.
OpenAI Five and AlphaStar from DeepMind used reinforcement learning to master Dota 2 and StarCraft II.
Waymo and Tesla use reinforcement learning in autonomous-driving stacks for decision-making.