DACE Research Group: Current Projects

1. Simulator-free variational inference schemes for neural stochastic differential equations with applications to state estimation, discovery of governing equations, reduced-complexity modeling, and generative modeling of time-series data
2. Scalable Gaussian process regression of parametrized spatial and spatio-temporal fields
3. Khatri-Rao neural operators
4. Low-rank deep neural decomposition architectures for learning from data
5. Preconditioning schemes for generative modeling with differential equations
6. Multimodal generative modeling algorithms with applications to neuroscience