coming next!
By joining our mailing list, calendar events will automatically appear and be updated on your calendar - no need to add them manually.
Guangyuan Weng
PhD Candidate at Northeastern University
October 19, 2026 at 10:00 AM PT -> my local time
Beyond Distance: Mobility Neural Embeddings Reveal Visible and Invisible Barriers in Urban Space poster
Abstract
Human mobility in cities is constrained by both visible structures (e.g., highways, rivers) and less visible social and institutional boundaries, yet identifying these barriers at scale remains an open challenge. Here, we analyze 25.4 million anonymized trajectories across 11 U.S. cities, covering ~90 million people, to measure a functional distance between neighborhoods that reflects observed movement rather than geographic proximity, using neural embeddings to extend this measurement to the many neighborhood pairs with sparse direct flow. This distance reveals mobility barriers---pairs of neighborhoods that are physically close but behaviorally distant. Across cities, scales, and time, the strongest predictors of these barriers are differences in amenities, administrative borders, and residential segregation by race and socioeconomic status, while physical infrastructure plays a more limited role. Barrier crossing is stratified, occurring more often outside commuting hours and among residents of diverse, higher-income, or transit-oriented areas. Our approach provides a scalable framework to detect and monitor mobility barriers shaping urban accessibility.
Speaker Bio
Guangyuan Weng is a Ph.D. candidate in Computer Science at Northeastern University, advised by Prof. Esteban Moro at the Network Science Institute. His research lies at the intersection of machine learning and computational social science, studying what shapes access to opportunity in cities. He uses large-scale mobility data and neural embeddings to measure the social and institutional barriers that divide urban space, and studies how large language models and recommender systems shape which places people see. More broadly, he is interested in responsible generative AI and representation learning for human behavior. He has also spent summers at Chewy and Wormpex AI Research, building multimodal AI systems for marketing video understanding and long-tail visual recognition. He received Northeastern's Outstanding PhD Award for Teaching in 2026. Before Northeastern, he was proud to be among ShanghaiTech's early undergraduate classes, doing research in computer vision and robotics.
Other Upcoming Seminars
Kehang Zhu Postdoctoral Fellow at Stanford Digital Economy Lab
Topic: Automated Social Science: Language Models as Scientist and Subjects
Siddharth Patwardhan Research Assistant Professor at Kellogg School of Management Northwestern University
Title: Scaling and Structural Transitions in Optimal Urban Transportation Systems
Tianming Liu Schmidt AI in Science Fellow at Michigan Institute of Data Science & AI in Society University of Michigan
Title: Human-Aligned LLM Agents for Behavior Modeling in Transportation
Minho Kim Postdoctoral Scholar at Stanford Urban Resilience Initiative
Title: Wildfire risk management in an era of climate extremes
Wytze de Vries PhD Canditate at Eindhoven University of Technology
Title: Competitor-aware Race Management for Electric Endurance Racing