Variational inference for generalized linear mixed models using partially noncentered parametrizations
From MaRDI portal
Abstract: The effects of different parametrizations on the convergence of Bayesian computational algorithms for hierarchical models are well explored. Techniques such as centering, noncentering and partial noncentering can be used to accelerate convergence in MCMC and EM algorithms but are still not well studied for variational Bayes (VB) methods. As a fast deterministic approach to posterior approximation, VB is attracting increasing interest due to its suitability for large high-dimensional data. Use of different parametrizations for VB has not only computational but also statistical implications, as different parametrizations are associated with different factorized posterior approximations. We examine the use of partially noncentered parametrizations in VB for generalized linear mixed models (GLMMs). Our paper makes four contributions. First, we show how to implement an algorithm called nonconjugate variational message passing for GLMMs. Second, we show that the partially noncentered parametrization can adapt to the quantity of information in the data and determine a parametrization close to optimal. Third, we show that partial noncentering can accelerate convergence and produce more accurate posterior approximations than centering or noncentering. Finally, we demonstrate how the variational lower bound, produced as part of the computation, can be useful for model selection.
Recommendations
- A general framework for the parametrization of hierarchical models
- A stochastic variational framework for fitting and diagnosing generalized linear mixed models
- Variational inference in nonconjugate models
- Concentration of tempered posteriors and of their variational approximations
- Conditionally structured variational Gaussian approximation with importance weights
Cites work
- scientific article; zbMATH DE number 6377992 (Why is no real title available?)
- scientific article; zbMATH DE number 192992 (Why is no real title available?)
- scientific article; zbMATH DE number 1222298 (Why is no real title available?)
- scientific article; zbMATH DE number 718799 (Why is no real title available?)
- scientific article; zbMATH DE number 1086057 (Why is no real title available?)
- scientific article; zbMATH DE number 1865509 (Why is no real title available?)
- scientific article; zbMATH DE number 1911984 (Why is no real title available?)
- A comparison of Bayesian and likelihood-based methods for fitting multilevel models
- A default conjugate prior for variance components in generalized linear mixed models (Comment on article by Browne and Draper)
- A general framework for the parametrization of hierarchical models
- A likelihood-based method for analysing longitudinal binary responses
- Approximate Inference in Generalized Linear Mixed Models
- Assessing the performance of variational methods for mixed logistic regression models
- Conditional Akaike information criterion for generalized linear mixed models
- Default Bayesian model determination methods for generalised linear mixed models
- Efficient parametrisations for normal linear mixed models
- Explaining variational approximations
- Fully simplified multivariate normal updates in non-conjugate variational message passing
- General design Bayesian generalized linear mixed models
- Mixed effects models and extensions in ecology with R
- Parameter Expansion for Data Augmentation
- Pattern recognition and machine learning.
- Seeking efficient data augmentation schemes via conditional and marginal augmentation
- Sensitivity analysis in Bayesian generalized linear mixed models for binary data
- Some Covariance Models for Longitudinal Count Data with Overdispersion
- Statistical Analysis of Financial Data in S-Plus
- Variational inference for large-scale models of discrete choice
- Variational message passing
Cited in
(16)- Conditionally structured variational Gaussian approximation with importance weights
- Fast Bayesian estimation of spatial count data models
- Variational inference for sparse spectrum Gaussian process regression
- A stochastic variational framework for fitting and diagnosing generalized linear mixed models
- Variational inferences for partially linear additive models with variable selection
- Streamlined mean field variational Bayes for longitudinal and multilevel data analysis
- Efficient data augmentation techniques for some classes of state space models
- On coregionalized multivariate Gaussian Markov random fields: construction, parameterization, and Bayesian estimation and inference
- On expectation propagation for generalised, linear and mixed models
- Variational inference for count response semiparametric regression
- Gaussian variational approximation with sparse precision matrices
- The limited-memory recursive variational Gaussian approximation (L-RVGA)
- Fast and accurate estimation of non-nested binomial hierarchical models using variational inference
- Stochastic variational inference for large-scale discrete choice models using adaptive batch sizes
- Bayesian Variational Inference for Exponential Random Graph Models
- Scalable Bayes via barycenter in Wasserstein space
This page was built for publication: Variational inference for generalized linear mixed models using partially noncentered parametrizations
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q252744)
