A Measure-Theoretic Variational Bayesian Algorithm for Large Dimensional Problems

DOI10.1137/140966575MaRDI QIDQ5174322zbMATH OpenOpenAlexFDO

Authors Aurélia Fraysse, Thomas Rodet

Publication date 17 February 2015

Published in SIAM Journal on Imaging Sciences (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1137/140966575

zbMATH Keywords

convergenceill-posed inverse problemsinfinite-dimensional convex optimizationiterative algorithmlarge-dimensional dictionary learning problemmetric spacenumerical examplessparse signal reconstructionvariational Bayesian methodology

Mathematics Subject Classification ID

Nonlinear ill-posed problems (47J06) Numerical solutions of ill-posed problems in abstract spaces; regularization (65J20) Numerical solution to inverse problems in abstract spaces (65J22) Convex programming (90C25) Signal theory (characterization, reconstruction, filtering, etc.) (94A12)

Recommendations

• Variational Bayesian approximation. A rigorous approach
• Mean-field variational approximate Bayesian inference for latent variable models
• Mean field variational Bayes for elaborate distributions
• Variational Bayesian inference with Gaussian-mixture approximations
• The computational asymptotics of Gaussian variational inference and the Laplace approximation
• Large Bayesian VARMAs
• Variational Bayes on manifolds
• A large-deviation principle for Dirichlet posteriors

Cited in

(4)

• The computational asymptotics of Gaussian variational inference and the Laplace approximation
• Approximate inference for constructing astronomical catalogs from images
• Efficient Variational Bayesian Approximation Method Based on Subspace Optimization
• Stochastic variational inference for large-scale discrete choice models using adaptive batch sizes

This page was built for publication: A Measure-Theoretic Variational Bayesian Algorithm for Large Dimensional Problems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5174322)