Very accurate posterior approximations based on finite mixtures of the hyperparameters conditionals.
Gustavo L. GilardoniPublished in: Comput. Stat. Data Anal. (2006)
Keyphrases
- hyperparameters
- finite mixtures
- posterior distribution
- bayesian framework
- gaussian process
- closed form
- em algorithm
- model selection
- mixture model
- cross validation
- expectation maximization
- bayesian inference
- minimum message length
- gaussian processes
- support vector
- maximum a posteriori
- random sampling
- maximum likelihood
- sample size
- bayesian methods
- prior information
- probabilistic model
- incremental learning
- noise level
- missing values
- image processing
- posterior probability
- parameter estimation
- latent variables
- incomplete data
- upper bound
- bayesian networks
- pairwise
- prior knowledge
- generative model
- markov random field