Machine learning for on-demand ride services: A survey

On-demand ride services have become an integral part of urban transportation systems and have notably influenced mobility patterns. This study provides a comprehensive review of the applications of machine learning (ML) techniques, including supervised learning, reinforcement learning, explainable AI, and generative AI, in the context of on-demand ride services. We investigate the role of ML in addressing two core challenges: supply-demand management, including demand forecasting, passenger choice modeling, and driver behavior analysis; and platform operations, such as dynamic pricing, matching, route planning, and vehicle repositioning. To promote reproducible research in the field, we also summarize public datasets that are crucial for benchmarking and validating ML-based solutions. By identifying key challenges such as latent demand estimation and Human-AI collaboration, this study aims to inspire interdisciplinary innovations for building a more sustainable and equitable urban transportation system.