Chen Mengyi, Doctor, National University of Singapore

唐贻发 研究员

Scalable Learning of Macroscopic Stochastic Dynamics

12月29号（周一）14:00-15:00，南楼733

Macroscopic dynamical descriptions of complex physical systems are crucial for understanding and controlling material behavior. With the growing availability of data and compute, machine learning has become a promising alternative to first-principles methods to build accurate macroscopic models from microscopic trajectory simulations. However, for spatially extended systems, direct simulations of sufficiently large microscopic systems that inform macroscopic behavior is prohibitive. In this talk, I will introduce a framework that learns the macroscopic dynamics of large microscopic systems using only small-system simulations. Our framework employs a partial evolution scheme to generate training data pairs by evolving large-system snapshots within local patches. We subsequently identify the closure variables associated with the macroscopic observables and learn the macroscopic dynamics using a custom loss. I will also briefly introduce our ongoing work to apply this framework to a realistic high-entropy alloy system.

Bio: Chen Mengyi is a PhD student in the Department of Mathematics at the National University of Singapore. She received her bachelor’s degree in Mathematics from the University of Chinese Academy of Sciences in 2023. Her research focuses on leveraging machine learning to study the macroscopic dynamics of complex systems. She has published work in top machine learning venues, including the Conference on Neural Information Processing Systems (NeurIPS). She is also an organizer of the ICLR 2025 Workshop on Machine Learning for Multiscale Processes.