Convergence rate of a new non-parametric change-point estimator within a linear model
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
We consider data initially derived from a homoscedastic linear model that undergoes a change at an unknown point in time, such as a change in model (not necessarily linear) or heteroscedasticity. Based on an idea of the sum of the squares of the residuals penalized introduced in Bardet ( 2025), we propose a new version of this criterion for estimating the change point. In the very general nonparametric framework we have adopted, we demonstrate that the convergence rate of this estimator is the same as in the parametric case to within a logarithmic term. To do this, we use concentration inequalities for sub-exponential variables.
Fields of Study & Subjects
Advanced Statistical Process Monitoring
Statistical Methods and Inference
Advanced Statistical Methods and Models
Publication details
- Journal / source
- HAL
- Publication date
- 1 January 2026
- Language
- EN
- Document type
- Preprint