Model-Based Offline Reinforcement Learning With Uncertainty Estimation and Policy Constraint

Keyphrases

• Uncertainty Estimation 100%
• Policy Constraints 100%
• Learning under Uncertainty 100%
• Offline Reinforcement Learning 100%
• Model Bias 28%
• Bounded States 28%
• Ensemble Network 28%
• Effective Method 14%
• Dropout 14%
• Adverse Effects 14%
• Bootstrap 14%
• Network Method 14%
• Reinforcement Learning Model 14%
• Reinforcement Learning Algorithm 14%
• Constraint Method 14%
• Policy Optimization 14%
• Distribution Shift 14%
• Bias Problem 14%
• Maximum Mean Discrepancy 14%
• Policy Learning 14%
• Monte Carlo Dropout 14%
• MuJoCo 14%

Computer Science

• Reinforcement Learning 100%
• Uncertainty Estimation 100%
• Effective Method 16%
• Experimental Result 16%
• Optimization Policy 16%
• Generalizability 16%
• Overestimation 16%
• Model-Based Reinforcement Learning 16%

Neuroscience

• Reinforcement Learning 100%
• Adverse Effect 14%