Change Point Inference for Non-Euclidean Data Sequences using Distance Profiles

We introduce a scan statistic for detecting and locating changes in sequences whose elements lie in a separable metric space. The method is tuning-parameter-free and fully nonparametric, with asymptotic guarantees and applications to distributional data, graph Laplacians, U.S. electricity-generation compositions, and Bluetooth proximity networks.

Recommended citation: Paromita Dubey and Minxing Zheng. (2023; revised 2026). "Change Point Inference for Non-Euclidean Data Sequences using Distance Profiles." arXiv:2311.16025.
Download Paper