Vector-valued Gaussian Processes on Riemannian Manifolds via Gauge Equivariant Projected Kernels.
Michael HutchinsonAlexander TereninViacheslav BorovitskiySo TakaoYee Whye TehMarc Peter DeisenrothPublished in: CoRR (2021)
Keyphrases
- vector valued
- gaussian processes
- riemannian manifolds
- reproducing kernel hilbert space
- gaussian process
- matrix valued
- euclidean space
- regression model
- hyperparameters
- bayesian framework
- multi task learning
- model selection
- semi supervised
- kernel methods
- special case
- kernel function
- loss function
- geometric structure
- data dependent
- cross validation
- latent variables
- learning theory
- vector space
- distance measure
- bayesian inference
- metric space
- feature extraction
- text classification
- supervised learning
- kernel matrix
- active learning
- feature space
- support vector
- positive definite
- valued data
- training data