Conformal changepoint localization

2026-02-05

AI Summary

This study addresses the problem of constructing finite-sample confidence sets for offline single change-point localization without distributional assumptions. Leveraging data exchangeability, the authors propose CONCH, a general conformal change-point localization method that establishes a conformal inference framework guaranteeing finite-sample coverage without relying on parametric models or asymptotic approximations. The core contributions include a conformal analogue of the Neyman–Pearson lemma, from which an optimal scoring function is derived, and a universality result showing that all distribution-free methods fall within the CONCH framework. Under mild assumptions, the normalized length of the resulting confidence sets converges to zero, enabling the method to produce sharp and valid confidence sets even for complex data such as images and text.

Technology Category

• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Machine Learning: Calibration & Uncertainty Quantification
• Search and Optimization: Non-convex Optimization

Application Category

• Security and Privacy: Large-scale security measurements
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

We study the problem of offline changepoint localization in a distribution-free setting. One observes a vector of data with a single changepoint, assuming that the data before and after the changepoint are iid (or more generally exchangeable) from arbitrary and unknown distributions. The goal is to produce a finite-sample confidence set for the index at which the change occurs without making any other assumptions. Existing methods often rely on parametric assumptions, tail conditions, or asymptotic approximations, or only produce point estimates. In contrast, our distribution-free algorithm, CONformal CHangepoint localization (CONCH), only leverages exchangeability arguments to construct confidence sets with finite sample coverage. By proving a conformal Neyman-Pearson lemma, we derive principled score functions that yield informative (small) sets. Moreover, with such score functions, the normalized length of the confidence set shrinks to zero under weak assumptions. We also establish a universality result showing that any distribution-free changepoint localization method must be an instance of CONCH. Experiments suggest that CONCH delivers precise confidence sets even in challenging settings involving images or text.

Problem

• changepoint localization
• distribution-free
• confidence set
• finite-sample coverage
• exchangeability

Innovation

• conformal inference
• changepoint localization
• distribution-free
• finite-sample coverage
• exchangeability