Marginalizing Gaussian process hyperparameters using sequential Monte Carlo.
Andreas SvenssonJohan DahlinThomas B. SchönPublished in: CAMSAP (2015)
Keyphrases
- gaussian process
- hyperparameters
- posterior distribution
- model selection
- cross validation
- bayesian inference
- bayesian framework
- gaussian processes
- closed form
- support vector
- random sampling
- noise level
- maximum a posteriori
- em algorithm
- maximum likelihood
- incremental learning
- gaussian process regression
- approximate inference
- prior information
- sample size
- incomplete data
- bayesian methods
- parameter space
- covariance function
- particle filter
- visual tracking
- particle filtering
- probabilistic model
- machine learning
- parameter settings
- missing values
- image reconstruction
- parameter estimation
- active learning
- prior knowledge
- decision trees
- image processing
- computer vision