Scientific Machine Learning is an emerging research area focused on the opportunities and challenges of machine learning in the context of complex applications across science, engineering, and medicine. The Center for Scientific Machine Learning is addressing these challenges through the development of new methods that weave together the perspectives of the field of Computational Science and Engineering — perspectives grounded in structured physics-based modeling where enforcing the governing physical laws brings the power to constrain an otherwise intractable solution space — and the perspectives of data-driven machine learning. The group's research projects bring together a diverse range of computing theories and algorithms, including large-scale optimization, inverse theory, reduced-order modeling, uncertainty quantification, Bayesian inference, optimal experimental design, data assimilation, physics-informed deep learning, interpretable machine learning, reinforcement learning, and high performance computing.

Directors

Uncertainty Quantification
Scientific Machine Learning
Inverse Problems

Scientific Machine Learning
High-Performance Computing

Faculty and Research Staff

Machine Learning
Optimization

Imaging
Computational Mechanics
High-Performance Computing

Machine Learning
Optimization

Computational Mechanics
Partial Differential Equations

Uncertainty Quantification
Inverse Problems
Optimization

Uncertainty Quantification
Computational Geosciences
Scientific Machine Learning

Applied Mathematics
Data Science

Numerical Analysis
Data Science
Scientific Computing

Machine Learning
Numerical Analysis
Scientific Computing
Partial Differential Equations

Uncertainty Quantification
Optimization
Data Science

Scientific Machine Learning
Computational Engineering

Students

Staff

Members outside the Oden Institute

Rochan Avlur, Benjamin Beal, Ke Chen, Krishanunni Giri, Juncai He, Conrad Li, Lewis Liu, Qijia Jiang, Hai Nguyen, Minh Nguyen, Yorick Sanders, Josh Taylor

Machine Learning for Physics-based Systems: Optimal Approximations, Architectures, and Training. AFSOR MURI.

Addressing the challenges of using machine learning to advance modeling and simulation capabilities in physical systems.

PI: K. Willcox, O. Ghattas | Collaborators: Jorge Nocedal (Northwestern), Hayden Schaeffer (CMU) and Stephen Wright (UW Madison) | Funding: AFSOR

Collaborative Research: Frameworks: Convergence of Bayesian inverse methods and scientific machine learning in Earth system models through universal differentiable programming.

PI: Patrick Heimbach | Funding: NSF

Intelligent Machine Learning for Real-Time Processing of Hyperspectral Video Streams

PI: Chandrajit Bajaj | Funding: Army Future Command

MuSiKAL: Multiphysics Simulations and Knowledge discovery through AI/ML technologies

PI: Clint Dawson | Collaborators: Ruby Leung (PNNL), Hartmut Kaiser (LSU), Joannes Westerink (Notre Dame) | Funding: DOE Office of Science

Learning with confidence: bootstrapping error estimates for stochastic optimization.

This project aims to develop theory and algorithm for uncertainty analysis and confidence bounds for nonconvex stochastic optimization algorithms. (PIs: R. Ward, P. Sarkar (UT SDS) )

PI: R. Ward, P. Sarkar (UT SDS) | Funding: NSF

Overcoming line-of-sight constraints

This project develops deep learning algorithms that optimize sensor placements in complicated environments subject to line-of-sight constraints.

PI: Richard Tsai

Verifiable, Control-Oriented Learning On The Fly. AFOSR MURI.

Develop mathematical foundations for on-the-fly learning of dynamical systems from limited data.

PI: U. Topcu, R. Ward, A. Israel (UT), C. Fefferman (Princeton), C. Rowley (Princeton), A. Ahmadi (Princeton), M. Snazier (Northeaster). | Collaborators: Hayden Schaeffer (CMU), Giang Tran (Waterloo), Holger Rauhut (Aachen). | Funding: AFSOR

Harnessing Self-Organizing Maps for the Discovery of Star Formation in Molecular Clouds.

PI: Stella Offner | Funding: NSF

RISE of the Machines: Robust, Interpretable, Scalable, Efficient Decision Support Department of Energy, AI for Complex Systems and Decision Support

Developing a new class of intelligent automation and decision-support methodologies that address the RISE (robustness, interpretability, scalability, and efficiency) challenges.

PI: K. Willcox, O. Ghattas | Collaborators: B. van Bloemen Waanders, J. Jakeman, J. Hart (Sandia); L. Ruthotto (Emory). | Funding: DOE

Machine learning of MRI of the Pancreas in T1D and Application to Hispanic Populations

PI: Jack Virostko, Chandrajit Bajaj

Data Access and the ECCO Ocean and Ice State Estimate

PI: Patrick Heimbach | Funding: NASA

A hybrid four-dimensional variational data assimilation / scientific machine learning framework for coupled Arctic Ocean-sea ice model parameter calibration, state estimation, and nowcasting.

PI: Patrick Heimbach | Funding: ONR

Shape Optimization through Deep Reinforcement Learning

ECCO: Understanding Sea Level, Ice, and Earthâ€™s Climate

PI: Patrick Heimbach | Funding: NASA / JPL-Caltech subcontract

Optimized Design of Materials for Optical Computing

PI: Chandrajit Bajaj

Models and algorithms for optimal surveillance and exploration of complex environments

This project includes three main thrusts: (i) Development of non-myopic greedy algorithms and study their properties and efficiency; (ii) Development of Deep Learning models and data generation for learning the gain-of-information functions; (iii) development of mathematical understanding of U-net used the Deep Learning models used in the project.

PI: Richard Tsai

Randomized Algorithms for Solving Linear Systems.

This project aims to develop and disseminate randomized algorithms for solving linear systems of equations.

PI: G. Martinsson, R. Ward, J. Tropp (CalTech) and V. Rokhlin (Yale) | Collaborators: D. Giannakis (Courant Institute, NYU) | Funding: NSF

Learning Optimal Aerodynamic Designs. ARPA-E Differentiate Program

Creating efficient, accurate, and scalable deep neural network (DNN) representations of solution of design optimization problems

PI: K. Willcox, O. Ghattas | Collaborators: J. Martins (Michigan) | Funding: ARPA

AEOLUS: Advances in Experimental Design, Optimal Control, and Learning for Uncertain Complex Systems

https://aeolus.oden.utexas.edu/. Developing a unified optimization-under-uncertainty framework for (1) learning predictive models from data and (2) optimizing experiments, processes, and designs, all in the context of complex, uncertain energy systems.

PI: O. Ghattas, K. Willcox, , G. Biros, M. Gunzburger, Robert Moser, J.T. Oden, C. Webster | Collaborators: F. Alexander (Brookhaven), J. Turner (Oak Ridge), E. Dougherty (TAMU), Y. Marzouk (MIT). | Funding: DOE

Multi-Fidelity Modeling of Rocket Combustor Dynamics

Advancing the state-of-the-art in reduced-order models and enable efficient prediction of instabilities in liquid fueled rocket combustion systems. Project page.

PI: K. Willcox | Collaborators: K. Duraisamy, C. Huang (Michigan); B. Peherstorfer (Courant). | Funding: AFSOR

Autonomous Aerial Cargo Operations at Scale

PI: Ufuk Topcu, co-PIs: John-Paul Clarke, Karen Willcox | Collaborators: Karen Marais, Defeng Sun (Purdue), Hamsa Balakrishnan (MIT), Allison Chang and Matthew Edwards (MIT Lincoln Laboratory), Willie Rockward (Morgan State University), Mel Davis (Cavan Solutions). | Funding: NASA University Leadership Initiative

Data-Driven Cyberphysical Systems

See project page for collaborators and more info.

PI: Ufuk Topcu | Funding: NSF

CAREER: Provably Correct Shared Control for Human-Embedded Autonomous Systems

See project page for more info.

PI: Ufuk Topcu | Funding: NSF