Nonparametric inference for ratios of densities via uniformly valid and powerful permutation tests

Alberto Bordino and Thomas B. Berrett · 2025

Read Nonparametric inference for ratios of densities via uniformly valid and powerful permutation tests by Alberto Bordino and Thomas B. Berrett (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-2505-24529.txt.

Introduction

Alberto Bordino and Thomas Berrett propose a new statistical test in 2025 for comparing how data spread. The test asks whether the ratio between two density curves matches a known shape, using samples from each. Ratios are key in transfer learning and causal work, where shift between data sets must be checked.

Their method shuffles pooled data with weights through a sampling scheme that keeps the test valid at any sample size. Under mild conditions the test reliably detects true gaps, and a kernel version hits the best possible detection rate in a stated sense. A variant learns the ratio from a spare sample first.

Catalog record

CREATORAlberto Bordino and Thomas B. Berrett
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE254 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-24529

Browse the full library · About the Tower of Babel