An objective prior for hyperparameters in normal hierarchical models.
James O. BergerDongchu SunChengyuan SongPublished in: J. Multivar. Anal. (2020)
Keyphrases
- hyperparameters
- hierarchical models
- maximum a posteriori
- prior information
- model selection
- hierarchical model
- cross validation
- support vector
- bayesian framework
- bayesian inference
- em algorithm
- closed form
- hierarchical structures
- maximum likelihood
- generative model
- random sampling
- learning algorithm
- gaussian process
- prior knowledge
- image reconstruction
- noise level
- incremental learning
- markov random field
- exact inference
- sample size
- expectation maximization
- training data
- incomplete data
- missing values
- hierarchical structure
- active learning
- multiscale