Exploration-Exploitation Tradeoff

• Exploration
  • increase knowledge by exposure to new experiences
  • gather information to make better decisions
  • best long-term strategy may involve short-term sacrifices
• Exploitation
  • exploiting known information to maximize the reward.
  • maximize performance based on current knowledge

The tradeoff

 balance how much we explore the environment and how much we exploit what we know about the environment.

Multi-Armed Bandit