Building relational world models for reinforcement learning

Mathematics

• Reinforcement Learning 100%
• Markov Decision Process 54%
• Reward 37%
• Model 21%
• Q-learning 18%
• Inductive Logic Programming 17%
• Value Iteration 15%
• Difference Method 12%
• Transition Probability 11%
• Policy 10%
• Estimate 10%
• Background 9%
• State Space 9%
• Knowledge 8%
• Evaluation 8%
• Performance 6%
• Standards 6%
• Form 4%

Engineering & Materials Science

• Markov processes 70%
• Reinforcement learning 66%
• Football 37%
• Inductive logic programming (ILP) 21%