A Matrix-free Approach for Solving the Parametric Gaussian Process Maximum Likelihood Problem.
Mihai AnitescuJie ChenLei WangPublished in: SIAM J. Sci. Comput. (2012)
Keyphrases
- gaussian process
- maximum likelihood
- hyperparameters
- likelihood function
- gaussian processes
- fully bayesian
- model selection
- bayesian framework
- gaussian process regression
- approximate inference
- regression model
- maximum a posteriori
- em algorithm
- gaussian process classification
- random sampling
- closed form
- latent variables
- parameter estimation
- gaussian process models
- semi supervised
- cross validation
- expectation maximization
- incremental learning
- support vector
- variational bayes
- sparse approximations
- gaussian mixture model
- expectation propagation
- covariance function
- higher order
- noise level
- data sets
- generative model
- machine learning