Scalable Bayesian computation for crossed and nested hierarchical models
From MaRDI portal
(Redirected from Publication:6184925)
Abstract: The article is about algorithms for learning Bayesian hierarchical models, the computational complexity of which scales linearly with the number of observations and the number of parameters in the model. It focuses on crossed random effect and nested multilevel models, which are used ubiquitously in applied sciences, and illustrates the methodology on two challenging real data analyses on predicting electoral results and real estate prices respectively. The posterior dependence in both classes is sparse: in crossed random effects models it resembles a random graph, whereas in nested multilevel models it is tree-structured. For each class we develop a framework for scalable computation. We provide a number of negative (for crossed) and positive (for nested) results for the scalability (or lack thereof) of methods based on sparse linear algebra, which are relevant also to Laplace approximation methods for such models. Our numerical experiments compare with off-the-shelf variational approximations and Hamiltonian Monte Carlo. Our theoretical results, although partial, are useful in suggesting interesting methodologies and lead to conclusions that our numerics suggest to hold well beyond the scope of the underlying assumptions.
Cites work
- A general framework for the parametrization of hierarchical models
- Analysis of variance -- why it is more important than ever. (With discussions and rejoinder)
- Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations (with discussion)
- Asymptotic Variance and Convergence Rates of Nearly-Periodic Markov Chain Monte Carlo Algorithms
- Auxiliary Gradient-Based Sampling Algorithms
- Backfitting for large scale crossed random effects regressions
- Bayesian Inference for Logistic Models Using Pólya–Gamma Latent Variables
- Efficient moment calculations for variance components in large unbalanced crossed random effects models
- Efficient parametrisations for normal linear mixed models
- Estimation and inference for very large linear mixed effects models
- Fast and accurate estimation of non-nested binomial hierarchical models using variational inference
- Gaussian Markov Random Fields
- Graphical models, exponential families, and variational inference
- Hierarchical array priors for ANOVA decompositions of cross-classified data
- scientific article; zbMATH DE number 1085989 (Why is no real title available?)
- scientific article; zbMATH DE number 1124118 (Why is no real title available?)
- Log-concave sampling: Metropolis-Hastings algorithms are fast
- Optimal Scaling of Discrete Approximations to Langevin Diffusions
- Parameter Expansion for Data Augmentation
- Pattern recognition and machine learning.
- Probabilistic Networks and Expert Systems
- Scalable inference for crossed random effects models
- Scalable logistic regression with crossed random effects
- Seeking efficient data augmentation schemes via conditional and marginal augmentation
- Spectral gap in random bipartite biregular graphs and applications
- Stability of the Gibbs sampler for Bayesian hierarchical models
- Streamlined Variational Inference for Linear Mixed Models with Crossed Random Effects
- Variance components.
Cited in
(7)- Greater than the sum of its parts: computationally flexible Bayesian hierarchical modeling
- Bayesian nonparametric models for multiple raters: a general statistical framework
- Partially factorized variational inference for high-dimensional mixed models
- Consistent and scalable composite likelihood estimation of probit models with crossed random effects
- Learning topic hierarchies by tree-directed latent variable models
- Conjugate Gradient Methods for High-Dimensional GLMMs
- Scalability of Metropolis-within-Gibbs schemes for high-dimensional Bayesian models
This page was built for publication: Scalable Bayesian computation for crossed and nested hierarchical models
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6184925)
