NIPS 2018 Expo Talk
Dec. 6, 2021
DL and RL in Intelligent Transportation Systems
Sponsor: Didi Chuxing
Jieping Ye (DiDi), Tony Qin (DiDi), Chenxi Wang (DiDi)
Jieping Ye (DiDi), Tony Qin (DiDi)
Data-enabled smart transportation has attracted a surge of interest from machine learning researchers nowadays due to the bloom of online ride-hailing industry and rapid development of autonomous driving. Large-scale high quality route data and trading data (spatiotemporal data) have been generated every day, which makes machine learning an urgent need and preferred solution for the decision making in intelligent transportation systems. While a large amount of work have been dedicated to traditional transportation problems, they are far from satisfactory for the rising need.
In this talk, we systematically introduce the basic services and the key emerging challenges in modern mobile transportation platform. We will introduce the deep learning and reinforcement learning solutions that we have adopted in our online map services. Meanwhile, Deep reinforcement learning (DRL) has achieved many successes in solving different types of sequential decision problems. We will walk through the evolution of DRL research at DiDi on order dispatching and driver repositioning. We hope it will inspire the audience and facilitate transportation AI research.