Student Seminar by Ms. Fangyi Ding on May 23, 2025 10AM

Title: A Graph Deep Learning Model for Station Ridership Prediction in Expanding Metro Networks

Speaker: Ms. Fangyi Ding (Department of Urban Planning and Design)

Date: May 23, 2025 (Friday)

Time: 10:00 am - 11:00 am

Venue: Room 612B, 6/F, Haking Wong Building, The University of Hong Kong

About the talk: Due to their reliability, efficiency, and environmental friendliness, metro systems have become a crucial solution to transportation challenges associated with urbanization. Many countries have constructed or expanded their metro networks over the past decades. During the planning stage, accurately predicting station ridership post-expansion, particularly for new stations, is essential to enhance the effectiveness of infrastructure investments. However, station-level metro ridership prediction under expansion scenarios (MRP-E) has not been thoroughly explored, as most advanced models currently focus on short-term predictions. MRP-E presents significant challenges due to the absence of historical data for newly built stations and the dynamic, complex spatiotemporal relationships between stations during expansion phases. In this study, we propose a Metro-specific Multi-Graph Attention Network model (Metro-MGAT) to address these issues. Our model leverages multi-sourced urban context data and network topology information to generate station features. Multi-relation graphs are constructed to capture the spatial correlations between stations, and an attention mechanism is employed to facilitate graph encoding. The model has been evaluated through realistic experiments using multi-year metro ridership data from Shanghai, China. The results validate the superior performance of our approach compared to existing methods, particularly in predicting ridership at new stations.

About the speaker: Fangyi Ding is currently a PhD candidate in the Department of Urban Planning and Design at the University of Hong Kong (HKU). She holds a master's degree from Tongji University and a bachelor's degree from the Harbin Institute of Technology. Her research interests lie in transportation demand forecasting, network modeling, and travel behavior analysis, with a particular focus on public transit and shared mobility services. Her work has been published in leading academic journals and conferences, including Transportation Research Part D, the Journal of Transport Geography, the Transportation Research Board (TRB) Annual Meeting, and ACM SIGSPATIAL, among others.

15 Apr 2025 (Tue) 10:30-12:00 | Seminar by Prof. Stein W. Wallace

Modeling with Stochastic Programming

Speaker:

Prof. Stein W. Wallace

Professor of Operational Research

Leader of the Centre for Shipping and Logistics

Centre for Shipping and Logistics, NHH Norwegian School of Economics

Date:    Apr 15, 2025 (Tuesday)

Time:   10:30 am – 12:00 nn

Venue:  Room 8-28, 8/F Haking Wong Building, The University of Hong Kong

Abstract

There are many deep papers on the mathematics and algorithmics of stochastic programming. But why should we, as operations research people interested in transportation and logistics, care? The world is stochastic for sure, but does that imply that we need stochastic models to get good decisions? And if we embark on a genuine application, where real money is involved, what are the modeling questions we need to pose? What are the steps we need to take before we arrive at mathematical and algorithmic challenges?

About the Speaker

Stein W. Wallace is a Professor of Operational Research and leader of the Centre for Shipping and Logistics at NHH Norwegian School of Economics. He received his Dr. Scient degree in informatics from the University of Bergen in 1984. He has earlier held professorships at for example Lancaster University Management School, The Chinese University of Hong Kong, Molde University College and NTNU, as well as visiting positions at for example Business School of Sichuan University in Chengdu, IBM Watson Research in NY, Columbia University, ENP Grenoble, France and The University of Washington. Wallace has published more than 120 papers in internationally leading journals such as Operations Research, Management Science, Production and Operations Management, Transportation Science, Transportation Research A, B and D and E, Mathematical Programming, European Journal of Operational Research, and INFORMS Journal on Computing. He is best known for his work in stochastic programming, in particular the two books Stochastic Programming (with Peter Kall from 1994) and Modeling with stochastic programming (with Alan King from 2012, Second Edition n 2024), but also for work in logistics and energy systems. He has over 14000 citations.

He has been on numerous editorial boards, including INFORMS Journal on Computing (1990-2018). He founded the Norwegian OR Society and has held elected positions in The British OR Society as well as The Society for Transportation and Logistics in INFORMS, The Mathematical Programming Society and EURO.

Student Seminar by Mr. Kaihang Zhang on Feb 27, 2025 4PM

Title: Tactical Operations of Service Region Dimensioning, Bundling, and Matching for On-Demand Food Delivery Services

Speaker: Mr. Kaihang Zhang (Department of Civil Engineering)

Date: Feb 27, 2025 (Thursday)

Time: 4:00 pm - 5:00 pm

Venue: Room 612B, 6/F, Haking Wong Building, The University of Hong Kong

About the talk: On-demand food delivery (OFD) services have experienced a significant surge in popularity in recent years, which poses various challenges for service operators. To address these challenges, we discuss an analytical model that captures the complex interplay of the OFD system by considering factors such as adjustable service region size and order bundling. We investigate how key decision variables, including maximum delivery distance and bundling ratio, affect the system's endogenous variables and two critical system performance metrics: customer total waiting time and order throughput. Our analysis yields several intriguing managerial insights. First, the maximum delivery distance has a non-monotonic impact on the customer accumulation time, delivery time, and total waiting time, and there is a “win-win” situation in which increasing the maximum delivery distance benefits both the customer total waiting time and order throughput. Second, order bundling is crucial under high customer demand to ensure adequate food delivery supply, but it is less desirable under low customer demand due to increased detour distances in delivery. We further explore strategies for minimizing customer total waiting time (by setting small service regions and bundling ratios) and order throughput (by establishing larger service regions). Recognizing the partial conflict between these two objectives, we identify the Pareto-efficient frontier that serves as a guideline for service operators in balancing these competing goals.

About the speaker: Kaihang Zhang is currently a PhD candidate at The University of Hong Kong supervised by Dr. Jintao Ke. He is a student member of INFORMS, ITS of HKU, and the recipient of Hong Kong PhD Fellowship and HKU Presidential Scholarship. He has been developing economic analytical models and network flow models throughout his PhD study at HKU on the operations for on-demand urban mobility systems. During his time at HKU, he worked as a visiting research student at LIMOS, University of Michigan, working on the development of data-driven analytical model for food delivery services. He has Bachelor’s Degrees from Zhejiang University and UIUC, and a Master’s Degree from UC Berkeley.