kcpRS-package: KCP on the running statistics

In kcpRS: Kernel Change Point Detection on the Running Statistics

KCP on the running statistics

Description

Flagging change points on a user-specified running statistics through KCP (Kernel Change Point) detection. A KCP permutation test is first implemented to confirm whether there is at least one change point (k>0) in the running statistics. If this permutation test is significant, a model selection procedure is implemented to choose the most optimal number of change points.

Details

This package contains the function kcpRS that can accept a user-defined function, RS_fun, which should derive the running statistics of interest. For examples, see runMean, runVar, runAR and runCorr. kcpRS performs a full change point analysis on the running statistics starting from locating the optimal change points given k, significance testing if k>0, and finally, determining the most optimal k. This function calls the function kcpa to find the most optimal change points given k and then the permTest function to carry out the permutation test. The model selection step is embedded in the kcpRS function.

This package also contains the function kcpRS_workflow which carries out a stepwise change point analysis to flag changes in 4 basic time series statistics: mean, variance, autocorrelation (lag 1) and correlations.

Two illustrative data sets are included: MentalLoad and CO2Inhalation

Author(s)

Jedelyn Cabrieto ([e-mail]) and Kristof Meers

For the core KCP analysis, the authors built upon the codes from the Supplementary Material available in doi:10.1080/01621459.2013.849605 by Matteson and James (2012).

References

Arlot, S., Celisse, A., & Harchaoui, Z. (2019). A kernel multiple change-point algorithm via model selection. Journal of Machine Learning Research, 20(162), 1-56.

Cabrieto, J., Tuerlinckx, F., Kuppens, P., Grassmann, M., & Ceulemans, E. (2017). Detecting correlation changes in multivariate time series: A comparison of four non-parametric change point detection methods. Behavior Research Methods, 49, 988-1005. doi:10.3758/s13428-016-0754-9

Cabrieto, J., Tuerlinckx, F., Kuppens, P., Hunyadi, B., & Ceulemans, E. (2018). Testing for the presence of correlation changes in a multivariate time series: A permutation based approach. Scientific Reports, 8, 769, 1-20. doi:10.1038/s41598-017-19067-2

Cabrieto, J., Tuerlinckx, F., Kuppens, P., Wilhelm, F., Liedlgruber, M., & Ceulemans, E. (2018). Capturing correlation changes by applying kernel change point detection on the running correlations. Information Sciences, 447, 117-139. doi:10.1016/j.ins.2018.03.010

Cabrieto, J., Adolf, J., Tuerlinckx, F., Kuppens, P., & Ceulemans, E. (2018). Detecting long-lived autodependency changes in a multivariate system via change point detection and regime switching models. Scientific Reports, 8, 15637, 1-15. doi:10.1038/s41598-018-33819-8

See Also

kcpRS

kcpRS_workflow

MentalLoad

CO2Inhalation

kcpRS documentation built on Oct. 25, 2023, 5:07 p.m.