Detecting multiple generalized change-points by isolating single ones

Media / Events / Detecting multiple generalized change-points by isolating single ones
17 September. 2020, 11:00, Online Webinar

Invited talk by Andreas Anastasiou

"Detecting multiple generalized change-points by isolating single ones"

17 September 2020 at 11am on ZOOM

In this talk, 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 in the trend, accompanied by discontinuities or not, in the piecewise-linear model. The number of change-points can increase with the sample size. The method is based on an isolation technique, which prevents the consideration of intervals that contain more than one changepoint. This isolation enhances ID’s accuracy as it allows for detection in the presence of frequent changes of possibly small magnitudes. Thresholding and model selection through an information criterion are the two stopping rules described in the talk. A hybrid of both criteria leads to a general method with very good practical performance. We show that ID is at least as accurate as the state-of-the-art methods; most of the times it outperforms them. The R package IDetect implementing the method described in the talk is available from CRAN.
Andreas_Anastasiou_ucy.jpg Andreas Anastasiou finished his D.Phil. from the Department of Statistics at the University of Oxford under the supervision of Professor Gesine Reinert. His D.Phil. research project was on finding explicit upper bounds on the distributional distance between the distribution of the Maximum Likelihood Estimator (MLE) and its approximate normal distribution. After his D.Phil. studies, he took a Postdoctoral Research Officer position at the Department of Statistics at the London School of Economics and Political Science. He has been working with Professor Piotr Fryzlewicz on new challenges in Time Series analysis and especially on developing new methods for change-point detection under complex structures. At the moment, Andreas is a Lecturer at the Department of Mathematics and Statistics at the University of Cyprus.


Subscribe to our Newsletter

* indicates required