Detecting multiple generalized change-points by isolating single ones

Abstract

We introduce a new approach, called Isolate-Detect (ID), for the consistent estimation of the number and location of multiple generalized change-points in noisy data sequences. Examples of signal changes that ID can deal with are changes in the mean of a piecewise-constant signal and changes, continuous or not, in the linear trend. The number of change-points can increase with the sample size. Our method is based on an isolation technique, which prevents the consideration of intervals that contain more than one change-point. This isolation enhances ID’s accuracy as it allows for detection in the presence of frequent changes of possibly small magnitudes. In ID, model selection is carried out via thresholding, or an information criterion, or SDLL, or a hybrid involving the former two. The hybrid model selection leads to a general method with very good practical performance and minimal parameter choice. In the scenarios tested, ID is at least as accurate as the state-of-the-art methods; most of the times it outperforms them. ID is implemented in the R packages IDetect and breakfast, available from CRAN.

Modeling of Covid-19 Pandemic in Cyprus

“…Even though binary segmentation is conceptually simple, it has the disadvantage that at each step of the algorithm, it looks for a single change-point, which leads to its suboptimality in terms of accuracy, especially for signals with frequent change-points. One method that works towards solving this issue is the Isolate-Detect (ID) methodology of Anastasiou and Fryzlewicz (2019); it is the method used for the analysis carried out in this paper.…”

“…In this work, we are using the Isolate-Detect (ID) methodology of Anastasiou and Fryzlewicz (2019) to detect changes based on (1) by using linear and constant signals, as described above; see Appendix A-1 for a description of the method.…”

The Impact of Government Actions and Risk Perception on the Promotion of Self-Protective Behaviors During the COVID-19 Pandemic

“…The time series analysis allowed us to identify a relatively stable evolution in the follow-up of the public health authorities' recommendations. However, although in general terms a high level of adherence was observed, the change-point detection analysis, using the Isolate-Detect methodology that we used previously 19, allowed us, to identify certain points that could indicate some abrupt changes in terms of adherence with the recommended behaviors (Fig. 2).…”

“…When it comes to forecasting, the role of the nal homogeneous segment (the data after the last change-point) is very important because it allows for a more accurate prediction of the future values of the data sequence at hand. The Isolate-Detect (ID) methodology 19 is employed in order to detect changes based on the model given in Eq. ( 1).…”

Narrowest Significance Pursuit: inference for multiple change-points in linear models

“…Gaussian noise with σ = 10, simulated with random seed set to 1. This represents a difficult setting from the perspective of multiple change-point detection, with practically all state of the art multiple change-point detection methods failing to estimate all 11 change-points with high probability (Anastasiou and Fryzlewicz, 2020). Therefore, a high degree of uncertainty with regards to the existence and locations of change-points can be expected here.…”