AMS Weekly Seminar | Yao Xie

Location: Shaffer 3

When: March 26th at 1:30 p.m.

Title: Matching-Based Flow Methods Beyond Data Imitation

Abstract: Modern generative models, including diffusion and flow-based methods, are often introduced as tools for data imitation. In this talk, I present a different viewpoint: matching-based flow methods can also be used as scalable solvers for optimization and game problems posed directly over probability distributions. I will discuss several examples in which the target distribution is not specified by the data alone but rather by an underlying variational or equilibrium principle, including worst-case generation in robust learning, dynamic optimal transport, and high-dimensional mean-field games. The key idea is to parameterize probability-space dynamics using deep flow models and to train them with matching objectives that avoid repeated simulation, thereby enabling practical methods in high dimensions. I will also discuss recent theoretical results on the convergence of the resulting iterative algorithms in probability space. Together, these examples suggest that matching-based generative flows are useful not only for sampling but also as a computational framework for high-dimensional optimization, game theory, and statistical inference at the level of distributions. 

Bio: Yao Xie is the Coca-Cola Foundation Chair and Professor in the H. Milton Stewart School of Industrial and Systems Engineering at the Georgia Institute of Technology, where she also serves as Associate Director of the Machine Learning Center (ML@GT). She received her Ph.D. in Electrical Engineering, with a minor in Mathematics, from Stanford University. Her research lies at the intersection of statistics, machine learning, and optimization, with a focus on statistically sound and computationally efficient methods for high-dimensional, sequential, and spatio-temporal data. She is a member of the 2026 cohort of the National Academies’ New Voices in Sciences, Engineering, and Medicine program and the IEEE Information Theory Society Distinguished Lecturer for 2026 to 2027. Her honors include the NSF CAREER Award, the INFORMS Gaver Early Career Award for Excellence in Operations Research, and the C.W.S. Woodroofe Award. She serves as an Associate Editor for multiple journals, including Operations Research, IEEE Transactions on Information Theory, and the Journal of the American Statistical Association (Theory and Methods), and as an Area Chair for NeurIPS, ICML, and ICLR.

Zoom link: https://wse.zoom.us/j/92366532431