Model Selection in Local Approximation Gaussian Processes: A Markov Random Fields Approach.
Hamed JalaliMartin PawelczykGjergji KasneciPublished in: IEEE BigData (2021)
Keyphrases
- gaussian processes
- markov random field
- model selection
- hyperparameters
- gaussian process
- marginal likelihood
- maximum a posteriori
- parameter estimation
- expectation propagation
- cross validation
- belief propagation
- random fields
- higher order
- graph cuts
- regression model
- bayesian learning
- closed form
- image segmentation
- bayesian methods
- sample size
- posterior distribution
- energy function
- machine learning
- bayesian framework
- approximate inference
- conditional random fields
- pairwise
- mixture model
- information criterion
- random sampling
- feature selection
- message passing
- em algorithm
- maximum likelihood
- free energy
- bayesian inference
- image reconstruction