Variational empirical Bayes variable selection in high-dimensional logistic regression

Yiqi Tang, and, Ryan Martin · 2025

Read Variational empirical Bayes variable selection in high-dimensional logistic regression by Yiqi Tang, and, Ryan Martin (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-10532.txt.

Description

Yiqi Tang and Ryan Martin speed up Bayesian variable selection for yes-no outcomes in a 2025 paper. Logistic regression with thousands of candidate predictors is common in genetics and medicine. Bayesian methods pick the real ones well but compute too slowly at that scale.

The team starts from an empirical Bayes method known for strong guarantees. Their variational shortcut stands in for the chance spread over which variables belong, not for the coefficients themselves. That choice keeps the math workable at that size. Simulations show the picking stays strong. The fast shortcut is proven to keep the first method's habit of finding the true ones. Practitioners get sound picking without the heavy cost that usually comes with it.

Catalog record

CREATORYiqi Tang, and, Ryan Martin
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE46 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-10532

Browse the full library · About the Tower of Babel