Scholarlink

The ridesharing system is a complex multi-agent system with multiple decision levers. RL offers a powerful modeling vehicle for optimizing this system, but as we have seen from the current literature, challenges remain in tackling complexity in the learning algorithms, the coordination among the agents, and the joint optimization of multiple levers. Along tackling these challenges, we expect that domain knowledge in ridesharing as well as transportation in general will be increasingly instrumental to the successful adoption of RL.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Closing Remarks

  • 章节

  • 发表日期：

    2025年

查看全文
  • Zhiwei (Tony) Qin,
  • Xiaocheng Tang,
  • Qingyang Li,
  • Hongtu Zhu,
  • Jieping Ye

摘要

The ridesharing system is a complex multi-agent system with multiple decision levers. RL offers a powerful modeling vehicle for optimizing this system, but as we have seen from the current literature, challenges remain in tackling complexity in the learning algorithms, the coordination among the agents, and the joint optimization of multiple levers. Along tackling these challenges, we expect that domain knowledge in ridesharing as well as transportation in general will be increasingly instrumental to the successful adoption of RL.