Machine Learning, Dynamical Systems and Control

Stacks Image 120

Physics Informed Machine Learning Workshop

University of Washington, Seattle
June 6-7, 2019

Local Organizers
Steven L. Brunton (Mechanical Engineering)
J. Nathan Kutz (Applied Mathematics)

Plenary Speakers

Stacks Image 16

Michael Mahoney, UC Berkeley
Why Deep Learning Works: Heavy-Tailed Random Matrix Theory as an Example of Physics Informed Machine Learning


Stacks Image 14

Michael Brenner, Harvard
Machine Learning for Partial Differential Equations
Stacks Image 24

Emily Fox, Washington
Flexibility, Interpretability, and Scalability in Time Series Modeling
Stacks Image 32

Charbel Farhat, Stanford
Probabilistic Physics-Based Machine Learning for Digital Twins

Minisymposia Sessions

Control and Optimization

Stacks Image 78

Sasha Aravkin, Washington
Algorithms for Nonsmooth, Nonconvex Problems in Data-Driven Discovery
Stacks Image 76

Benjamin Erichson, Berkeley
Shallow Learning for Flow Reconstruction with Limited Sensors and Limited Data
Stacks Image 62

Alex Gorodetsky, Michigan
Scalable Learning of Dynamical Systems

Data-Driven Model Discovery

Stacks Image 87

Steven Brunton, Washington
Discovering interpretable and generalizable dynamical systems from data
Stacks Image 92

Kathleen Champion, Washington
Data-driven discovery of coordinates and governing equations
Stacks Image 97

Paris Perdikaris, Pennsylvania
Data-driven modeling of stochastic systems using physics-aware deep learning


Reduced Order Modeling

Stacks Image 105

Kevin Carlberg, Sandia
Breaking Komolgorov-Width Barriers using Deep Learning: Projection Dynamical Systems on Nonlinear Manifolds Constructed by Convolutional Autoencoders
Stacks Image 110

Karthik Duraisamy, Michigan
Physics constrained probabilistic learning of Koopman decompositions
Stacks Image 115

Benjamin Peherstorfer, Courant NYU
Data generation and time-delay corrections for learning reduced models with operator inference