Intuitive joint priors for Bayesian linear multilevel models: the R2D2M2 prior
From MaRDI portal
Abstract: The training of high-dimensional regression models on comparably sparse data is an important yet complicated topic, especially when there are many more model parameters than observations in the data. From a Bayesian perspective, inference in such cases can be achieved with the help of shrinkage prior distributions, at least for generalized linear models. However, real-world data usually possess multilevel structures, such as repeated measurements or natural groupings of individuals, which existing shrinkage priors are not built to deal with. We generalize and extend one of these priors, the R2D2 prior by Zhang et al. (2020), to linear multilevel models leading to what we call the R2D2M2 prior. The proposed prior enables both local and global shrinkage of the model parameters. It comes with interpretable hyperparameters, which we show to be intrinsically related to vital properties of the prior, such as rates of concentration around the origin, tail behavior, and amount of shrinkage the prior exerts. We offer guidelines on how to select the prior's hyperparameters by deriving shrinkage factors and measuring the effective number of non-zero model coefficients. Hence, the user can readily evaluate and interpret the amount of shrinkage implied by a specific choice of hyperparameters. Finally, we perform extensive experiments on simulated and real data, showing that our inference procedure for the prior is well calibrated, has desirable global and local regularization properties and enables the reliable and interpretable estimation of much more complex Bayesian multilevel models than was previously possible.
Cites work
- A comparison of Bayesian and likelihood-based methods for fitting multilevel models
- A survey of Bayesian predictive methods for model assessment, selection and comparison
- Adaptive Bayesian density estimation with location-scale mixtures
- Applied asymptotic analysis
- Bayesian and frequentist regression methods
- Bayesian Regression Using a Prior on the Model Fit: The R2-D2 Shrinkage Prior
- Constructing Priors that Penalize the Complexity of Gaussian Random Fields
- Default Bayesian analysis with global-local shrinkage priors
- Dirichlet-Laplace priors for optimal shrinkage
- Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models
- Generalized inverse normal distributions
- Generating generalized inverse Gaussian random variates by fast inversion
- Graphical test for discrete uniformity and its applications in goodness-of-fit evaluation and multiple sample comparison
- Handbook of Markov Chain Monte Carlo
- scientific article; zbMATH DE number 720689 (Why is no real title available?)
- scientific article; zbMATH DE number 6734253 (Why is no real title available?)
- Kumaraswamy's distribution: a beta-type distribution with some tractability advantages
- Large-scale multiple hypothesis testing with the normal-beta prime prior
- Logistic-Normal Distributions: Some Properties and Uses
- On the half-Cauchy prior for a global scale parameter
- Penalising model component complexity: a principled, practical approach to constructing priors
- Prior distributions for variance parameters in hierarchical models (Comment on article by Browne and Draper)
- Projective inference in high-dimensional problems: prediction and feature selection
- Regression and other stories
- Ridge Regression: Biased Estimation for Nonorthogonal Problems
- Shrinkage priors for Bayesian penalized regression
- Sparsity information and regularization in the horseshoe and other shrinkage priors
- The Bayesian Lasso
- The horseshoe estimator for sparse signals
- The no-U-turn sampler: adaptively setting path lengths in Hamiltonian Monte Carlo
Cited in
(10)- Some models are useful, but how do we know which ones? Towards a unified Bayesian model taxonomy
- Efficient estimation and correction of selection-induced bias with order statistics
- Spatial regression modeling via the R2D2 framework
- A standardization procedure to incorporate variance partitioning-based priors in latent Gaussian models
- Design and structure dependent priors for scale parameters in latent Gaussian models
- The group R2D2 shrinkage prior for sparse linear models with grouped covariates
- Simulation-based calibration checking for Bayesian computation: the choice of test quantities shapes sensitivity
- Advances in projection predictive inference
- The ARR2 prior: flexible predictive prior definition for Bayesian auto-regressions
- Generalized decomposition priors on R2
This page was built for publication: Intuitive joint priors for Bayesian linear multilevel models: the R2D2M2 prior
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6170612)
