Choosing the Right Norm for Change Point Detection in Functional Data
Patrick Bastian · 2025
Read Choosing the Right Norm for Change Point Detection in Functional Data by Patrick Bastian (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-04476.txt.
Description
Patrick Bastian, writing alone in 2025, argues for a different measuring stick when hunting change points in curve data. Many data sets arrive as whole curves over time, one per subject or day. A change point is where the average curve suddenly shifts. Past tests measured that shift with two common norms. Bastian proposes a test based on the L1 norm, which sums absolute gaps. He proves the test is valid for classic and for broader hypotheses.
Theory and experiments both favor the new stick in many settings. Its edge is widest across a broad range of cases, he shows. A power booster is added for cases where the shift is sparse. The method is shown on made-up data and on real data as well. Choosing the right norm turns out to matter a great deal. The paper is a methods piece for statisticians who work with curve data.
Catalog record
| CREATOR | Patrick Bastian |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 59 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-04476