This article explores how Gaode leverages AI and large‑scale spatio‑temporal pre‑training to transform uncertain urban transit into reliable, real‑time bus and subway predictions, detailing technical challenges, model architectures, and user‑focused features that enhance daily commuting.
JD Tech showcased a remarkable 21-paper presence at AAAI 2021, covering federated learning, spatio‑temporal AI, recommendation systems, computer vision, and causal learning, highlighting the company’s transition from research to real‑world AI applications across smart cities, retail, and risk management.
JD Digits announced that 16 of its research papers were accepted at the prestigious AAAI 2021 conference, covering federated learning, vertical federated learning, communication‑efficient SGD, spatio‑temporal graph diffusion for traffic forecasting, robust meta‑learning for sales prediction, graph‑enhanced session recommendation, knowledge‑aware social recommendation, and causal learning for retail delinquency, highlighting the company's strong AI research and its real‑world smart‑city and industry applications.
