Bayesian variable selection in high dimensional problems without assumptions on prior model probabilities

TL;DR

This paper addresses high-dimensional linear model variable selection without relying on assumptions about prior model probabilities, enabling effective selection when the number of regressors exceeds the sample size.

Contribution

It introduces a novel Bayesian variable selection method that operates without assumptions on prior model probabilities in high-dimensional settings.

Findings

Method performs well even when p exceeds n

Effective in small sample scenarios

Outperforms existing approaches in simulations

Abstract

We consider the problem of variable selection in linear models when $p$, the number of potential regressors, may exceed (and perhaps substantially) the sample size $n$ (which is possibly small).