Skip to content

Publications

Selected published work and a preprint. STAMP, currently under review, is described on its project page.

T-STAR: A Context-Aware Transformer Framework for Short-Term Probabilistic Demand Forecasting in Dock-Based Shared Micro-Mobility

Jingyi Cheng, Gonçalo Correia, Oded Cats, Shadi Sharif Azadeh

Transportation Research Part C: Emerging Technologies 2026

Two-stage probabilistic forecasting for sparse demand, evaluated across 235 bike-sharing stations with zero-shot transfer experiments.

DOICode

A Short-Term Predict-Then-Cluster Framework for Meal Delivery Services

Jingyi Cheng, Shadi Sharif Azadeh

Data Science for Transportation 2025

Connecting point and distributional demand forecasts with spatially constrained clustering for operational planning.

DOICode

Real-Time Integrated Dispatching and Idle Fleet Steering with Deep Reinforcement Learning for A Meal Delivery Platform

Jingyi Cheng, Shadi Sharif Azadeh

Preprint · arXiv:2501.05808 2025

A dual-control deep reinforcement learning framework connecting predicted demand with courier dispatching and steering.

Read preprint