Machine Learning Applications

Page in progress. This research area page is being developed. Check back soon, or return to the Research overview to explore other areas.

The Machine Learning Applications thrust develops and applies machine learning techniques to accelerate engineering analysis and design. Key directions include:

Neural PDE solvers for parametric and irregular geometries
NURBS-Diff: differentiable programming with Non-Uniform Rational B-Splines
ML-accelerated topology optimization for high-resolution, manufacturable designs
Neural implicit geometry representations for flow and mechanics simulations
Multi-level CNNs for 3D object recognition and classification