We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.ML

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Machine Learning

Title: Approximation Properties of Variational Bayes for Vector Autoregressions

Abstract: Variational Bayes (VB) is a recent approximate method for Bayesian inference. It has the merit of being a fast and scalable alternative to Markov Chain Monte Carlo (MCMC) but its approximation error is often unknown. In this paper, we derive the approximation error of VB in terms of mean, mode, variance, predictive density and KL divergence for the linear Gaussian multi-equation regression. Our results indicate that VB approximates the posterior mean perfectly. Factors affecting the magnitude of underestimation in posterior variance and mode are revealed. Importantly, We demonstrate that VB estimates predictive densities accurately.
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG); Econometrics (econ.EM); Computation (stat.CO)
Cite as: arXiv:1903.00617 [stat.ML]
  (or arXiv:1903.00617v1 [stat.ML] for this version)

Submission history

From: Reza Hajargasht [view email]
[v1] Sat, 2 Mar 2019 03:24:16 GMT (45kb)

Link back to: arXiv, form interface, contact.