Matérn Gaussian processes on Riemannian manifolds

Abstract

Gaussian processes are an effective model class for learning unknown functions, particularly in settings where accurately representing predictive uncertainty is of key importance. Motivated by applications in the physical sciences, the widely-used Matérn class of Gaussian processes has recently been generalized to model functions whose domains are Riemannian manifolds, by re-expressing said processes as solutions of stochastic partial differential equations. In this work, we propose techniques for computing the kernels of these processes on compact Riemannian manifolds via spectral theory of the Laplace-Beltrami operator in a fully constructive manner, thereby allowing them to be trained via standard scalable techniques such as inducing point methods. We also extend the generalization from the Matérn to the widely-used squared exponential Gaussian process. By allowing Riemannian Matérn Gaussian processes to be trained using well-understood techniques, our work enables their use in mini-batch, online, and non-conjugate settings, and makes them more accessible to machine learning practitioners.

References in corpus (4)

• Practical Bayesian Optimization of Machine Learning Algorithms
• Gaussian Processes for Data-Efficient Learning in Robotics and Control
• Numerical solution of fractional elliptic stochastic PDEs with spatial white noise
• Bayesian Optimization Meets Riemannian Manifolds in Robot Learning

Cited by in corpus (12)

• Unpacking the Expressed Consequences of AI Research in Broader Impact Statements
• Equivalence of measures and asymptotically optimal linear prediction for Gaussian random fields with fractional-order covariance operators
• Nonnegative spatial factorization
• The SPDE approach for Gaussian and non-Gaussian fields: 10 years and still running
• Scalable Thompson Sampling using Sparse Gaussian Process Models
• Pathwise Conditioning of Gaussian Processes
• Isometric Gaussian Process Latent Variable Model for Dissimilarity Data
• Multi-group Gaussian Processes
• Matérn Gaussian Processes on Graphs
• Non-separable Spatio-temporal Graph Kernels via SPDEs
• Scalable Variational Gaussian Processes via Harmonic Kernel Decomposition
• Uniform Generalization Bounds for Overparameterized Neural Networks