From: 2025-06-10 09:00 to 10:00
Place: E:2116, Dept. of Computer Science, E-building, Klas Anshelms väg 10/Ole Römers väg 3, Lund
Contact: emma [dot] soderberg [at] cs [dot] lth [dot] se

LTH AI and Digitalization Breakfast Seminar: Bayesian Learning for Compact Dynamical Representations of Nonlinear Systems with Mengwu Guo

Time: Wednesday June 10th, 09:00-10:00 (come early to grab breakfast before the presentation starting 09:15)

Location: E:2116, E-building, Ole Römers väg 3

Sign-up link: https://forms.gle/QqkqGonGPhmsKbiNA

Speaker: Mengwu Guo

Title: Bayesian Learning for Compact Dynamical Representations of Nonlinear Systems

Abstract: Credible real-time simulation is a crucial enabler for digital twin technology, and data-driven model reduction is a key approach to achieving it. In this talk, we will discuss non-intrusive Bayesian methods for learning reduced-order representations of high-dimensional dynamical systems, with built-in probabilistic quantification of modeling uncertainties to certify computational reliability. The core strategy involves using Bayesian inference for the parametrization inspired by projection-based model reduction. Particularly, Gaussian process approximations are leveraged to formulate differential-equation-constrained likelihood functions and hence improve predictive performance, especially when training data are noisy and/or scarce. These techniques have demonstrated their effectiveness in data-driven reduced-order modeling by delivering accurate temporal predictions along with robust uncertainty quantification.

Bio: Mengwu Guo is an Associate Professor in Numerical Analysis and Scientific Computing at the Centre for Mathematical Sciences, Lund University, Sweden. Before joining Lund, he served as an Assistant Professor in the Department of Applied Mathematics at the University of Twente, the Netherlands. He previously held postdoctoral positions at the Oden Institute for Computational Engineering and Sciences at the University of Texas at Austin (USA) and at the Institute of Mathematics at EPFL (Switzerland). Mengwu earned both his bachelor's and doctoral degrees in Civil Engineering, with top honors, from Tsinghua University, China. His research spans computational science and numerical mathematics, with recent emphasis on data-driven model reduction, uncertainty quantification, and physics-aware probabilistic machine learning.

Organizer: LTH Profile Area: AI and Digitalization

(https://www.lth.se/english/research/profile-areas/pillars-of-ai-and-digitalization/)