Problem
Research questions and friction points this paper is trying to address.
multiple change-point detection
interval localization
false discovery rate control
multiple testing
Interval Localization for Multiple Change-points with Error Rate Control
🤖 AI Summary
This study addresses the challenge of controlling the false discovery rate (FDR) in interval localization for multiple changepoint detection. It establishes an algorithm-agnostic framework that decouples the detection and selection stages via permutation testing, thereby achieving strict finite-sample FDR control. Furthermore, this work proposes the POIS algorithm and its enhanced variant, POIS+, which leverage order-preserving sample splitting and e-value weighting techniques to transform detection evidence into optimal weights. Theoretically, the proposed methods guarantee FDR control while substantially improving statistical power. Extensive simulation studies and real-data applications validate both the efficiency and practical utility of the approach.
📝 Abstract
Interval localization in multiple change-point detection aims to identify intervals containing true change-points while controlling false discoveries. We formulate this task as a multiple testing problem over detector-reported intervals and develop an algorithm-agnostic and distribution-free framework that controls the interval-level FDR, the expected proportion of retained intervals containing no true change-points. The proposed permutation-based order-preserving interval selection (POIS) uses order-preserving sample splitting to separate detection from permutation testing. To improve power, POIS+ reuses the detection-side evidence and introduces detection-conditional $e$-values as unnormalized weights for permutation $p$-values. Both procedures achieve finite-sample FDR control. Under the stated conditions, POIS attains asymptotically full power while POIS+ has asymptotically no lower power than POIS. Simulation results show that POIS maintains high power and POIS+ further improves power with empirical FDR below the nominal level for both procedures. Two real data applications demonstrate the practical effectiveness of the proposed methods.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Multiple change-point detection
False discovery rate control
Permutation testing
Order-preserving sample splitting
e-values
🔎 Similar Papers
No similar papers found.
💼 Related Jobs
No related jobs found.
