Offline changepoint localization using a matrix of conformal p-values

TL;DR

This paper introduces the MCP algorithm that uses conformal p-values to accurately locate changepoints in data sequences, providing confidence intervals under mild assumptions, and demonstrates its effectiveness on diverse datasets.

Contribution

The paper presents a novel MCP algorithm leveraging conformal p-values for changepoint detection with theoretical guarantees and practical classifier-based implementation.

Findings

Effective changepoint localization on synthetic data

Successful application to image, text, and accelerometer data

Theoretical validation via a new conformal Neyman-Pearson lemma

Abstract

Changepoint localization is the problem of estimating the index at which a change occurred in the data generating distribution of an ordered list of data, or declaring that no change occurred. We present the broadly applicable MCP algorithm, which uses a matrix of conformal p-values to produce a confidence interval for a (single) changepoint under the mild assumption that the pre-change and post-change distributions are each exchangeable. We prove a novel conformal Neyman-Pearson lemma, motivating practical classifier-based choices for our conformal score function. Finally, we exemplify the MCP algorithm on a variety of synthetic and real-world datasets, including using black-box pre-trained classifiers to detect changes in sequences of images, text, and accelerometer data.