Loading Events

« All Events

  • This event has passed.

PICS Colloquium: Information Theory of Multiscale Simulations with Daniel Tartakovsky

October 25 @ 2:00 PM3:00 PM

Speaker: Dr. Tartakovsky is a Professor in Energy Science and Engineering Department, Institute for Computational Mathematics and Engineering, and Bio-X at Stanford University. Prior to joining Stanford in 2015, he was a Staff Scientist in Theoretical Division at Los Alamos National Laboratory (1996-2005) and Professor of Mechanical and Aerospace Engineering at University of California San Diego (2005-2015). His research interests include scientific computing, data-driven modeling and simulation, parameter estimation, and uncertainty quantification, with applications ranging from electrochemical energy storage to biomedical engineering.

Abstract: We present an information-theoretic approach for integration of multi-resolution data into multiscale simulations. Fine-scale information can comprise observational data and/or simulation results related to both system states and system parameters. It is aggregated into its coarse-scale representation by setting a probabilistic equivalence between the two scales, with parameters that are determined via minimization of observables error and mutual information across scales. The same quantities facilitate the use of coarse-scale data to constrain compatible fine-scale distributions. In the second part of this talk, we leverage the information-geometric properties of the statistical manifold to reduce predictive uncertainty via data assimilation. Specifically, we exploit the information-geometric structures induced by two discrepancy metrics, the Kullback-Leibler divergence and the Wasserstein distance, which explicitly yield natural gradient descent. The use of a deep neural network as a surrogate model for MD enables automatic differentiation, further accelerating optimization. The manifold’s geometry is quantified without sampling, yielding an accurate approximation of the gradient descent direction. Our numerical experiments demonstrate that accounting for the manifold’s geometry significantly reduces the computational cost of data assimilation by both facilitating the calculation of gradients and reducing the number of required iterations.

 

Details

Date:
October 25
Time:
2:00 PM – 3:00 PM

Organizer

Delaney Parks
Phone
7034701288
Email
dkparks@seas.upenn.edu

Venue

Penn Institute for Computational Science
3401 Walnut Street, 5th Floor
Philadelphia, PA 19104 United States
+ Google Map