Note on a non-parametric method for change-point detection

TL;DR

This paper details R code for a non-parametric change-point detection method based on the Wilcoxon-Mann-Whitney test, validated through simulations, offering an alternative to existing methods like Pettitt's.

Contribution

It introduces a detailed implementation of a non-parametric change-point detection method and compares it to Pettitt's approach, validated via simulations.

Findings

Validated through simulations showing effectiveness

Provides R code for practical implementation

Competitor to Pettitt's method using Wilcoxon-Mann-Whitney test

Abstract

The purpose of this note is to present in details R codes to implement a non-parametric method for change-point detection. The proposed approach is validated from various perspectives using simulations. This method is a competitor to that of Pettitt ([3]) and is, like the latter, based on the Wilcoxon-Mann-Whitney test. It is used in [4] for the study of relatively short time series obtained from measurements on cores sampled in the bay of Brest.