Hybrid Multi-agent Deep Reinforcement Learning for Autonomous Mobility on Demand Systems

Keyphrases

• Autonomous Mobility-on-demand (AMoD) 100%
• Multi-agent Deep Reinforcement Learning (MADRL) 100%
• Action Space 50%
• Novel Combinations 50%
• Control Policy 50%
• Markov Decision Process 50%
• Experiment-based 50%
• Multi-agent 50%
• Sequential Decision Problems 50%
• Performance Stability 50%
• Taxi Data 50%
• Computational Tractability 50%
• Anticipatory Control 50%
• Soft Actor-critic 50%
• Coordinated Decision 50%
• Weighted Bipartite Matching 50%
• Request Assignment 50%
• Profit-maximizing 50%
• Rejection Decision 50%

Computer Science

• Markov Decision Process 100%
• multi agent 100%
• Multi-Agent Deep Reinforcement Learning 100%
• bipartite matching 100%
• Sequential Decision Making 100%