Score-based change point detection via tracking the best of infinitely many experts

Anna Markovich, Nikita Puchkin
8/26/2024
cs.LGstat.MEstat.ML

No official code found. Best implementation: lipperr/score_based_change_point_detection (0 stars, last commit 2026-02-04).

Abstract

We suggest a novel algorithm for online change point detection based on sequential score function estimation and tracking the best expert approach. The core of the procedure is a version of the fixed share forecaster for the case of infinite number of experts and quadratic loss functions. The algorithm shows a promising performance in numerical experiments on artificial and real-world data sets. We also derive new upper bounds on the dynamic regret of the fixed share forecaster with varying parameter, which are of independent interest.

View on arXivView PDF

Code Implementations(1)

lipperr/score_based_change_point_detectionCommunity100%
00PythonMay 2, 20248 months ago

Similar Papers

Greedy online change point detection

82% match2023

Sequential Change Point Detection via Denoising Score Matching

82% match1 repo12025

The Cost of Learning under Multiple Change Points

81% match2026

On Multi-entity, Multivariate Quickest Change Point Detection

80% match2025

Online Heavy-tailed Change-point detection

80% match2023

Robust and Scalable Bayesian Online Changepoint Detection

79% match1 repo242023

Cite this paper

@article{markovich2024scorebased,
  title  = {Score-based change point detection via tracking the best of infinitely many experts},
  author = {Anna Markovich and Nikita Puchkin},
  year   = {2024},
  eprint = {2408.14073},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2408.14073v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2408.14073.svg)](https://codeofpaper.com/paper/2408.14073)

Discussion