Visit ShanghaiTech University | 中文 | How to find us
HOME > News and Events > Events
Bayesian Deep Learning for Predicting Modeling in Science and Engineering
Date: 2018/7/18             Browse: 55

Speaker:     Prof. Nicholas Zabaras. ND

Time:          10:15—11:15, July 18 

Location:    Room 1A-200, SIST Building

Host:          Prof. Qifeng Liao

Abstract:

We will briefly review recent advances in the solution of stochastic PDEs using Bayesian deep encoder-decoder networks. These models have been shown to work remarkably well for uncertainty quantification tasks in very-high dimensions. In this talk through examples in computational physics and chemistry, we will address their potential impact for (a) modeling dynamic problems (b) accounting for model form uncertainty in coarse grained simulations and (c) providing the means to coarse graining and automatic collective variable selection in atomistic models. We will start with an application to dynamic multiphase flow problems typical of geological carbon storage process-based multiphase flow models. We will show the ability to capture dynamics as well as discontinuities in both the spatial and stochastic domain. We will continue by investigating the potential of4 improving the Reynolds Average Navier Stokes Equations based turbulence models by replacing the anisotropic Reynolds stress with a LES data trained Bayesian neural network. Challenges in capturing physical invariance relations in the design of the neural network will be outlined. Finally, we will discuss the use of encoder-decoder networks in automatically coarse graining atomistic systems and in deriving a generative model of the high-dimensional atomic configuration space. We formulate the discovery of collective variables (CVs) as a Bayesian inference problem and consider the CVs as hidden generators of the full-atomistic trajectory. This allows us to compute estimates of observables as well as our probabilistic confidence on them. The discovered CVs are related to physicochemical properties which are essential for understanding mechanisms especially in unexplored complex systems.

Bio: 

Prof. Nicholas Zabaras joined Notre Dame in 2016 as the Viola D. Hank Professor of Computational Science and Engineering after serving as Uncertainty Quantification Chair and founding director of the “Warwick Centre for Predictive Modeling (WCPM)” at the University of Warwick. He is the Director of the interdisciplinary University of Notre Dame “Center for Informatics and Computational Science (CICS)” that aims to bridge the areas of data-sciences, scientific computing and uncertainty quantification for complex multiscale/multiphysics problems in science and engineering. He is also serving as the Hans Fisher Senior Fellow with the Institute for Advanced Study at the Technical University of Munich where recently was appointed "TUM Ambassador". He is also an Honorary Professor at the Dept. of Mathematics at the University of Hong Kong. Prior to this, he spent 23 years serving in all academic ranks of the faculty at Cornell University where he was the director of the “Materials Process Design and Control Laboratory (MPDC)”. He received his Ph.D. in Theoretical and Applied Mechanics from Cornell, after which he started his academic career at the faculty of the University of Minnesota. Professor Zabaras' research focuses on the integration of computational mathematics, statistics, and scientific computing for the predictive modeling of complex systems. He has been honored with the Wolfson Research Merit Award from the Royal Society, the Michael Tien '72 Excellence in Teaching Prize from Cornell University, and the Presidential Young Investigator Award from the National Science Foundation.

SIST-Seminar 18065