Vector-valued Gaussian Processes on Riemannian Manifolds via Gauge Independent Projected Kernels
This work addresses the need for deploying Gaussian processes in physical science applications involving non-Euclidean domains, representing an incremental advancement in extending existing models to handle vector fields on manifolds.
The authors tackled the problem of modeling vector fields on Riemannian manifolds, such as spheres and tori, by proposing a method to construct gauge-independent kernels that enable vector-valued Gaussian processes, extending standard training techniques like variational inference for practical use.
Gaussian processes are machine learning models capable of learning unknown functions in a way that represents uncertainty, thereby facilitating construction of optimal decision-making systems. Motivated by a desire to deploy Gaussian processes in novel areas of science, a rapidly-growing line of research has focused on constructively extending these models to handle non-Euclidean domains, including Riemannian manifolds, such as spheres and tori. We propose techniques that generalize this class to model vector fields on Riemannian manifolds, which are important in a number of application areas in the physical sciences. To do so, we present a general recipe for constructing gauge independent kernels, which induce Gaussian vector fields, i.e. vector-valued Gaussian processes coherent with geometry, from scalar-valued Riemannian kernels. We extend standard Gaussian process training methods, such as variational inference, to this setting. This enables vector-valued Gaussian processes on Riemannian manifolds to be trained using standard methods and makes them accessible to machine learning practitioners.