Replacing Rewards with Examples: Example-Based Policy Search via Recursive Classification

Keyphrases

• Reward Function 100%
• Example-based 100%
• Policy Search 100%
• Reinforcement Learning Algorithm 50%
• Successful Outcome 50%
• Value Function 16%
• Control Algorithm 16%
• Bellman Equation 16%
• Debugging 16%
• Hyperparameters 16%
• Similar Problem 16%
• Lines of Code 16%
• Technical Expertise 16%
• Transition Outcomes 16%

Computer Science

• Reinforcement Learning 100%
• Control Algorithm 33%
• Lines of Code 33%
• Leaning Parameter 33%
• Probability 33%