A runtime monitoring framework to enforce invariants on reinforcement learning agents exploring complex environments

Mathematics

• Reinforcement Learning 100%
• Monitoring 68%
• Framework 38%
• Invariant 37%
• Monitor 28%
• Prior Knowledge 19%
• Safety 18%
• Learning 15%
• Learning Environment 14%
• Trial and error 11%
• Context 10%
• Robotics 10%
• Reward 9%
• Machine Learning 9%
• Training 8%
• Policy 8%

Engineering & Materials Science

• Reinforcement learning 66%
• Monitoring 37%
• Software agents 27%
• Robotics 16%
• Machine learning 16%
• Feedback 14%