Well-Posed Bayesian Inverse Problems with Infinitely Divisible and Heavy-Tailed Prior Measures.
Bamdad HosseiniPublished in: SIAM/ASA J. Uncertain. Quantification (2017)
Keyphrases
- heavy tailed
- inverse problems
- prior distribution
- image reconstruction
- maximum a posteriori
- posterior distribution
- global optimization
- convex optimization
- bayesian framework
- bayesian inference
- generalized gaussian
- prior information
- probability distribution
- bayesian methods
- optimization methods
- probability density function
- optimization problems
- bayesian networks
- generative model
- conditional probabilities
- maximum likelihood
- partial differential equations
- prior knowledge
- prior model
- computer vision
- multiresolution
- early vision
- wavelet transform
- hyperparameters
- image denoising
- high dimensional
- video sequences
- parameter estimation