NTU-CEE Distinguished Speaker Series: Professor Yafeng YIN

Organized By

CEE Seminar Committee

Host By

Associate Professor WANG Zhiwei

Topic

Abstract

Civil infrastructure systems are socio-technical networks whose performance depends critically on human decision-making. Choices made by users, operators, and other stakeholders shape system dynamics, resource allocation, resilience, and the effectiveness of policy interventions. Yet choice modeling remains a persistent bottleneck because high-quality choice data are expensive to collect, difficult to scale, and often limited in their ability to capture heterogeneous responses to new conditions. Large language models (LLMs) create a new opportunity by generating synthetic responses in structured choice settings at scale. However, raw LLM outputs are not behaviorally calibrated and therefore cannot be used directly for structural choice estimation or policy analysis.

This talk presents a theoretically grounded framework for converting LLMs into calibrated choice simulators and integrating LLM-generated responses with observed human data to estimate policy-credible structural choice models. Transportation systems serve as the primary technical context for illustrating the framework and its empirical implementation, while the underlying ideas are broadly applicable across civil infrastructure systems. The framework preserves interpretability and valid inference while leveraging generative AI to enrich choice data and improve behavioral realism. The central methodological question is how to align LLM-generated choices with human heterogeneity and substitution patterns, and then use this alignment to improve estimation, prediction, and policy evaluation.

Biography

Dr. Yafeng Yin is the Donald Cleveland Collegiate Professor of Engineering and the Donald Malloure Department Chair of Civil and Environmental Engineering at the University of Michigan, Ann Arbor. His research examines the interactions among travelers, infrastructure, and mobility services, aiming to enhance the design, operation, and regulation of transportation systems. He has published over 160 peer-reviewed papers in leading journals and served as Editor-in-Chief of Transportation Research Part C: Emerging Technologies from 2014 to 2020. He currently serves as an Area Editor for Transportation Science and an Associate Editor for Transportation Research Part B: Methodological, and co-Editor-Chief for the newly launched journal Artificial Intelligence for Transportation.

Dr. Yin’s work has received multiple honors, including the Frank M. Masters Transportation Engineering Award from American Society of Civil Engineers (ASCE), the Monroe-Brown Foundation Education Excellence Award from Michigan Engineering, the Doctoral Mentoring Award from the University of Florida, and the Outstanding Leadership Award from the Chinese Overseas Transportation Association (COTA). He has also received several best paper awards from the Transportation Research Board, including the Stella Dafermos, Ryuichi Kitamura, and Kikuchi-Karlaftis awards. He earned his Ph.D. from the University of Tokyo in 2002, and his master’s and bachelor’s degrees from Tsinghua University in 1996 and 1994, respectively.