PAC reinforcement learning with an imperfect model

Keyphrases

• Real Environment 100%
• Reinforcement Learning 100%
• State Action 100%
• Sample Complexity 100%
• Imperfect Models 100%
• Number of States 50%
• SIMPLE Algorithm 50%
• Single State 50%
• Approximate Model 50%
• Reinforcement Learning Algorithm 50%
• Natural Conditions 50%
• Reinforcement Learning Agent 50%
• Hard Examples 50%
• High-fidelity Simulator 50%
• Action Space 50%
• Complexity Guarantees 50%

Computer Science

• Reinforcement Learning 100%
• Simple Algorithm 33%
• Learning Agent 33%
• Natural Condition 33%
• Simulated Environment 33%
• Foundational Result 33%