Graph-based multiple change-point detection

TL;DR

This paper introduces a graph-based framework for detecting multiple change-points in multivariate and non-Euclidean data, combining segmentation algorithms with a new goodness-of-fit measure to improve accuracy and interpretability.

Contribution

The paper presents a novel change-point detection method that integrates graph-based statistics with segmentation and pruning techniques, outperforming existing approaches.

Findings

Outperforms existing change-point detection methods in various settings

Effectively detects change-points in complex multivariate data

Hierarchical arrangement of change-points enhances interpretability

Abstract

We propose a new multiple change-point detection framework for multivariate and non-Euclidean data. First, we combine graph-based statistics with wild binary segmentation or seeded binary segmentation to search for a pool of candidate change-points. We then prune the candidate change-points through a novel goodness-of-fit statistic. Numerical studies show that this new framework outperforms existing methods under a wide range of settings. The resulting change-points can further be arranged hierarchically based on the goodness-of-fit statistic. The new framework is illustrated on a Neuropixels recording of an awake mouse.