Differentially private scale testing via rank transformations and percentile modifications

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Levine, Joshua, Ramsay, Kelly
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866918083559424000
author Levine, Joshua
Ramsay, Kelly
author_facet Levine, Joshua
Ramsay, Kelly
contents We develop a class of differentially private two-sample scale tests, called the rank-transformed percentile-modified Siegel--Tukey tests, or RPST tests. These RPST tests are inspired both by recent differentially private extensions of some common rank tests and some older modifications to non-private rank tests. We present the asymptotic distribution of the RPST test statistic under the null hypothesis, under a very general condition on the rank transformation. We also prove RPST tests are differentially private, and that their type I error does not exceed the given level. We uncover that the growth rate of the rank transformation presents a tradeoff between power and sensitivity. We do extensive simulations to investigate the effects of the tuning parameters and compare to a general private testing framework. Lastly, we show that our techniques can also be used to improve the differentially private signed-rank test.
format Preprint
id arxiv_https___arxiv_org_abs_2507_03725
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Differentially private scale testing via rank transformations and percentile modifications
Levine, Joshua
Ramsay, Kelly
Methodology
Machine Learning
62G10, 62G20, 62G30, 68P27
G.3.1; G.3.5; K.6.5
We develop a class of differentially private two-sample scale tests, called the rank-transformed percentile-modified Siegel--Tukey tests, or RPST tests. These RPST tests are inspired both by recent differentially private extensions of some common rank tests and some older modifications to non-private rank tests. We present the asymptotic distribution of the RPST test statistic under the null hypothesis, under a very general condition on the rank transformation. We also prove RPST tests are differentially private, and that their type I error does not exceed the given level. We uncover that the growth rate of the rank transformation presents a tradeoff between power and sensitivity. We do extensive simulations to investigate the effects of the tuning parameters and compare to a general private testing framework. Lastly, we show that our techniques can also be used to improve the differentially private signed-rank test.
title Differentially private scale testing via rank transformations and percentile modifications
topic Methodology
Machine Learning
62G10, 62G20, 62G30, 68P27
G.3.1; G.3.5; K.6.5
url https://arxiv.org/abs/2507.03725