Surrogate modeling & model reduction
Cross-
Cutting
Research Area
Harnessing the data through the lens of physics-based modeling
Embracing the opportunities and challenges of machine learning in complex applications across science, engineering and medicine
Existing machine learning approaches do not have the robustness, reliability, scalability, or efficiency to make them viable for grand challenge problems in science and engineering.
An Overview: Scientific Machine Learning
What is Scientific Machine Learning?
Scientific Machine Learning brings together the complementary perspectives of computational science and computer science to craft a new generation of machine learning methods for complex applications across science and engineering. In these applications, dynamics are complex and multiscale, data are sparse and expensive to acquire, decisions have high consequence, and uncertainty quantification is essential. The greatest challenges facing society — clean energy, climate change, sustainable urban infrastructure, access to clean water, personalized medicine and more — by their very nature require predictions that go well beyond the available data. Scientific machine learning achieves this by incorporating the predictive power, interpretability and domain knowledge of physics-based models.
The applications are characterized by complex multiscale multiphysics dynamics, so that small changes in parameters can lead to large changes in system behavior.
The parameter space is very high dimensional. Many parameters of interest are fields (infinite dimensional). Without the constraints of physics, the solution space is so vast that driving decisions with data alone is doomed to failure.
Data are sparse and typically rely on physical sensing infrastructure, making them expensive to acquire. Data may be large in volume, but they provide only limited peeks into the underlying high-dimensional parameter space.
Uncertainty quantification of predictions must provide quantified confidence in the recommended decisions. This is especially challenging but especially important as we extrapolate beyond the data to issue predictions about future states.
Current research areas
Research is multifaceted, ranging from foundational advances in theory, methods and algorithms, to real-world impact in societal grand challenge problems.
Bayesian inverse problems
Physics-informed deep learning
Data assimilation
Interpretable machine learning
Reinforcement learning
Digital twins
Optimal experimental design
Working with partners
The Oden Institute and The Alan Turing Institute have a memorandum of understanding to collaborate in the areas of artificial intelligence for science and engineering, computational science and engineering, scientific machine learning, and data-centric engineering. Established in 2015, The Alan Turing Institute is a high-profile, vibrant and multidisciplinary national institute, bringing together 13 leading universities from England and Scotland, making it the UK’s national institute for data science and artificial intelligence.
The Oden Institute's Center for Scientific Machine Learning has strong collaborations with Department of Energy programs, including the AEOLUS Multifaceted Mathematics Integrated Capability Center, the ARPA-E DIFFERENTIATE program for design intelligence, and the Artificial Intelligence and Decision Support for Complex Systems program. Our faculty also played a key role in the ASCR visioning report on Basic Research Needs for Scientific Machine Learning
.Centers and Groups
To learn more about projects and people in Scientific Machine Learning, explore the centers and groups with research activities in this cross-cutting research area.
Center for Scientific Machine Learning
Optimization, Inversion, Machine Learning, and Uncertainty for Complex Systems
Computational Research in Ice and Ocean Systems Group
Probabilistic and High Order Inference, Computation, Estimation, and Simulation
News in brief
News
Aug. 20, 2026
The Quest to Make Climate Models Faster, Smarter and More Accurate
In a recent paper a team of researchers, including Professor Patrick Heimbach and Ph.D. student Joseph Kump, introduced a new framework for simulating components of the Earth. To demonstrate the framework’s versatility, the team tested it on four different Earth component models including an idealized ocean model of the Drake Passage.
News
Aug. 17, 2026
New Faculty Nick Nelsen Joins UT Austin to Advance Trustworthy AI Research
News
May 13, 2026
Seeing the Possibilities: High School Students Step into Oden Institute Research World
Students from the Liberal Arts and Science Academy got a front row seat to current research on topics including cancer, the cosmos, digital twins, robotics, and ocean systems during a field trip to the Oden Institute.