Variable selection using shrinkage priors

DOI10.1016/J.CSDA.2016.10.008zbMATH Open1466.62135arXiv1503.04303OpenAlexW2104844823MaRDI QIDQ1658484FDOQ1658484

Authors: Hanning Li, Debdeep Pati 

Publication date: 14 August 2018

Published in: Computational Statistics and Data Analysis (Search for Journal in Brave)

Abstract: Variable selection has received widespread attention over the last decade as we routinely encounter high-throughput datasets in complex biological and environment research. Most Bayesian variable selection methods are restricted to mixture priors having separate components for characterizing the signal and the noise. However, such priors encounter computational issues in high dimensions. This has motivated continuous shrinkage priors, resembling the two-component priors facilitating computation and interpretability. While such priors are widely used for estimating high-dimensional sparse vectors, selecting a subset of variables remains a daunting task. In this article, we propose a general approach for variable selection with shrinkage priors. The presence of very few tuning parameters makes our method attractive in comparison to adhoc thresholding approaches. The applicability of the approach is not limited to continuous shrinkage priors, but can be used along with any shrinkage prior. Theoretical properties for near-collinear design matrices are investigated and the method is shown to have good performance in a wide range of synthetic data examples.

Full work available at URL: https://arxiv.org/abs/1503.04303

Recommendations

• Variable Selection with Shrinkage Priors via Sparse Posterior Summaries
• Bayesian variable selection with shrinking and diffusing priors
• Bayesian variable selection for mixed effects model with shrinkage prior
• Variable Selection and Shrinkage: Comparison of Some Approaches
• Shrinkage Inverse Regression Estimation for Model-Free Variable Selection
• Bayesian variable selection
• A variational Bayes approach to variable selection
• scientific article; zbMATH DE number 1034042
• On Bayesian lasso variable selection and the specification of the shrinkage parameter

zbMATH Keywords

horseshoeMarkov chain Monte Carlovariable selectionBayesianshrinkage priors

Mathematics Subject Classification ID

Computational methods for problems pertaining to statistics (62-08)

Cites Work

• The Adaptive Lasso and Its Oracle Properties
• Bayesian Model Assessment and Comparison Using Cross-Validation Predictive Densities
• Inference with normal-gamma prior distributions in regression problems
• Sparse Bayesian infinite factor models
• The horseshoe estimator for sparse signals
• The Bayesian Lasso
• Sure Independence Screening for Ultrahigh Dimensional Feature Space
• Optimal predictive model selection.
• A Selective Overview of Variable Selection in High Dimensional Feature Space (Invited Review Article)
• Bayesian Variable Selection in Linear Regression
• 10.1162/15324430152748236
• Title not available (Why is that?)
• Dirichlet-Laplace priors for optimal shrinkage
• The Intrinsic Bayes Factor for Model Selection and Prediction
• Title not available (Why is that?)
• Consistent high-dimensional Bayesian variable selection via penalized credible regions
• Title not available (Why is that?)
• On the use of Non-Local Prior Densities in Bayesian Hypothesis Tests
• Title not available (Why is that?)
• Bayesian Variable Selection Under Collinearity
• Decoupling shrinkage and selection in Bayesian linear models: a posterior summary perspective

Cited In (25)

• Structured Shrinkage Priors
• Neuronized Priors for Bayesian Sparse Linear Regression
• Joint Bayesian estimation of voxel activation and inter-regional connectivity in fMRI experiments
• Horseshoe Regularisation for Machine Learning in Complex and Deep Models¹
• Shrinkage priors for Bayesian penalized regression
• Nearly optimal Bayesian shrinkage for high-dimensional regression
• Bayesian regression with undirected network predictors with an application to brain connectome data
• Detecting communities in attributed networks through bi-direction penalized clustering and its application
• Bayesian variable selection for multioutcome models through shared shrinkage
• Bayesian group bridge for bi-level variable selection
• Shrinkage Inverse Regression Estimation for Model-Free Variable Selection
• Sparse linear mixed model selection via streamlined variational Bayes
• Bayesian fusion estimation via \(t\) shrinkage
• Decoupling shrinkage and selection in Bayesian linear models: a posterior summary perspective
• A comparative study on high-dimensional bayesian regression with binary predictors
• Bayesian variable selection for mixed effects model with shrinkage prior
• Bayesian Approaches to Shrinkage and Sparse Estimation
• Shrinkage and variable selection by polytopes
• Shrinkage priors via random imaginary data
• Bayesian variable selection with shrinking and diffusing priors
• The beta-mixture shrinkage prior for sparse covariances with near-minimax posterior convergence rate
• Variable selection via penalized credible regions with Dirichlet-Laplace global-local shrinkage priors
• Bayesian tensor response regression with an application to brain activation studies
• A global-local approach for detecting hotspots in multiple-response regression
• Bayesian hierarchical modeling on covariance valued data

This page was built for publication: Variable selection using shrinkage priors

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1658484)