Gaussian Process Uniform Error Bounds with Unknown Hyperparameters for Safety-Critical Applications.

Alexandre Capone Armin Lederer Sandra Hirche

Published in: ICML (2022)

Keyphrases

error bounds
gaussian process
hyperparameters
safety critical
model selection
cross validation
gaussian processes
bayesian framework
closed form
bayesian inference
random sampling
support vector
posterior distribution

prior information
sample size
noise level
formal methods
theoretical analysis
incremental learning
maximum likelihood
em algorithm
worst case
gaussian process regression
maximum a posteriori
incomplete data
marginal likelihood

variational bayes
covariance function
parameter space
parameter settings
regression model
machine learning
missing values
noise reduction
probabilistic model
image reconstruction
active learning
upper bound

supervised learning
markov random field
generative model
semi supervised
expectation maximization
lower bound
image segmentation
decision trees
image processing
expectation propagation
feature selection
noisy images