Ridge-Regularized Largest Root Test For High-Dimensional General Linear Hypotheses

Haoran Li · 2025

Read Ridge-Regularized Largest Root Test For High-Dimensional General Linear Hypotheses by Haoran Li (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-2504-15510.txt.

Description

Haoran Li, in a solo paper from 2025, proposes a stabilized version of a classic test for group differences when there are many variables. The classic largest root test is strong at finding concentrated signals, but it breaks down when data dimensions approach the sample size because its matrices become unstable.

Li adds a ridge term, a small stabilizing addition, to the covariance estimates behind the test. He then derives the long-run distribution of the largest eigenvalue of the stabilized matrix under weak assumptions. A fast procedure estimates the centering and scaling values the test needs. Power analysis covers low-rank alternatives, where the true signal lives in a few directions. Tests on brain imaging and behavior data illustrate the method.

Catalog record

CREATORHaoran Li
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE249 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-15510

Browse the full library · About the Tower of Babel