Learning with Maximum A-Posteriori Perturbation Models.
Andreea GaneTamir HazanTommi S. JaakkolaPublished in: AISTATS (2014)
Keyphrases
- maximum a posteriori
- bayesian framework
- prior knowledge
- markov random field
- image reconstruction
- maximum likelihood
- reinforcement learning
- wavelet transform
- markov networks
- em algorithm
- hidden variables
- generalized gaussian
- machine learning
- posterior distribution
- generative model
- graphical models
- similarity measure