Scalable Bayesian variable selection and model averaging under block orthogonal design

TL;DR

This paper introduces a scalable Bayesian variable selection and model averaging framework for linear models with block-diagonal Gram matrices, offering exact solutions for block designs and heuristic approximations for general designs, implemented in R.

Contribution

The paper presents a novel, efficient algorithm for Bayesian variable selection under block-diagonal assumptions and extends it to general designs via spectral clustering, improving scalability and accuracy.

Findings

Exact Bayesian model selection for block-diagonal designs without numerical integration.

Efficient computation of posterior probabilities using one-dimensional integrals.

Scalable algorithm with linear cost in number of blocks, suitable for parallel processing.

Abstract

We propose a scalable algorithmic framework for exact Bayesian variable selection and model averaging in linear models under the assumption that the Gram matrix is block-diagonal, and as a heuristic for exploring the model space for general designs. In block-diagonal designs our approach returns the most probable model of any given size without resorting to numerical integration. The algorithm also provides a novel and efficient solution to the frequentist best subset selection problem for block-diagonal designs. Posterior probabilities for any number of models are obtained by evaluating a single one-dimensional integral that can be computed upfront, and other quantities of interest such as variable inclusion probabilities and model averaged regression estimates by carrying out an adaptive, deterministic one-dimensional numerical integration. The overall computational cost scales linearly with the number of blocks, which can be processed in parallel, and exponentially with the block size, rendering it most adequate in situations where predictors are organized in many moderately-sized blocks. For general designs, we approximate the Gram matrix by a block-diagonal using spectral clustering and propose an iterative algorithm that capitalizes on the block-diagonal algorithms to explore efficiently the model space. All methods proposed in this article are implemented in the R library mombf.