An Optimal False Discovery Rate Controlling Procedure for Changepoint Detection

Louis Davis, Guenther Walther · 2026

Read An Optimal False Discovery Rate Controlling Procedure for Changepoint Detection by Louis Davis, Guenther Walther (2026) 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-2608-00219.txt.

Description

Davis and Walther propose a changepoint method that controls the false discovery rate. The task is detecting and localizing a growing number of changepoints. Their data are a sequence of independent observations. The method is called Lean Bonferroni Detection, or LBD-FDR. It yields localized regions where a changepoint likely sits.

The method guarantees false discovery control across wide settings. Non-parametric and heavy-tailed data are included, a hard case. For Gaussian sequences, they derive when all strong changepoints are detected consistently. The method attains the optimal detection constant in certain regimes. It can beat the minimax Type I method in power in some settings. A feasible algorithm and simulations against five rivals round out the work.

Catalog record

CREATORLouis Davis, Guenther Walther
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE127 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2608.00219

Browse the full library · About the Tower of Babel