Path-Following Methods for Maximum a Posteriori Estimators in Bayesian Hierarchical Models: How Estimates Depend on Hyperparameters.
Zilai SiYucong LiuAlexander StrangPublished in: SIAM J. Optim. (2024)
Keyphrases
- maximum a posteriori
- hyperparameters
- posterior distribution
- cross validation
- maximum likelihood
- bayesian inference
- bayesian methods
- bayesian framework
- model selection
- markov random field
- image reconstruction
- em algorithm
- hierarchical models
- expectation maximization
- closed form
- gaussian process
- noise level
- support vector
- incremental learning
- prior information
- parameter estimation
- edge detection
- prior knowledge
- markov networks
- pairwise
- hierarchical structures
- feature space