A Frequency-Domain NonStationarity Test for dependent data

M. Ould Haye, A. Philippe · 2025

Read A Frequency-Domain NonStationarity Test for dependent data by M. Ould Haye, A. Philippe (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-2510-24319.txt.

Description

Mohamedou Ould Haye and Anne Philippe give a new test to tell if a data series is steady in time or not. Two cases are hard to split. A series can have long memory, with slow decay of links over time. Or it can be unsteady, with its traits shifting. Both give slow decay that looks the same at a glance. Past tests fail on long memory data or misjudge the rate of false alarms near the split.

Their test uses no fixed constants. It compares the frequency plots of the series across a few time blocks. They work out the limit laws for both the steady and the unsteady case. Those laws are sums of weighted chi square terms. In trials, the new test did well next to past tests, with sound behavior near the hard split cases.

Catalog record

CREATORM. Ould Haye, A. Philippe
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE38 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-24319

Browse the full library · About the Tower of Babel