Gaussian Process regression over discrete probability measures: on the non-stationarity relation between Euclidean and Wasserstein Squared Exponential Kernels.
Antonio CandelieriAndrea PontiFrancesco ArchettiPublished in: CoRR (2022)
Keyphrases
- gaussian process regression
- probability measures
- gaussian processes
- euclidean distance
- gaussian process
- stochastic process
- hilbert space
- metric space
- euclidean space
- probability measure
- non stationary
- kernel methods
- reproducing kernel hilbert space
- finite number
- kernel function
- semi supervised
- bayesian framework
- machine learning
- regression model
- finite dimensional
- infinite dimensional
- feature space
- multiscale