ϕ-ML
Physics Informed machine learning See Physics informed statistical learning of battery lifetimes
Constrained Variational Inference
Approach to solving Bayesian inverse problems using constrained VI
- Variational autoencoding of PDE inverse problems
- Deep orthogonal decompositions for convective nowcasting
Hybrid Modelling of Dynamical Systems
Latent Force Models
A nonparametric model of a dynamic system is more than a GP with a temporal argument. One promising idea of embedding dynamic systems of embedding dynamical systems properties was introduced in Lawrence and Alvarez. My thesis considered an extension to allow for multiplicative interactions between the latent forces and state variables.
Multiplicative Latent Force Models
