Rebalancing Docked Bicycle Sharing System with Approximate Dynamic Programming and Reinforcement Learning

Business & Economics

• Approximate Dynamic Programming 100%
• Reinforcement Learning 85%
• Rebalancing 79%
• Markov Decision Process 52%
• Learning Dynamics 30%
• Computational Complexity 26%
• Urban Environment 26%
• Dynamic Systems 25%
• Buffer 23%
• Numerical Experiment 23%
• Dynamic Programming 22%
• Safety 20%
• Operator 20%
• Prediction 18%
• Alternatives 10%

Engineering & Materials Science

• Bicycles 78%
• Reinforcement learning 67%
• Dynamic programming 66%
• Markov processes 23%
• Pickups 13%
• Random forests 11%
• Dynamical systems 9%
• Computational complexity 9%
• Experiments 4%