Change-point detection in the covariance kernel of functional data using data depth

12/02/2021
by   Kelly Ramsay, et al.
0

We investigate several rank-based change-point procedures for the covariance operator in a sequence of observed functions, called FKWC change-point procedures. Our methods allow the user to test for one change-point, to test for an epidemic period, or to detect an unknown amount of change-points in the data. Our methodology combines functional data depth values with the traditional Kruskal Wallis test statistic. By taking this approach we have no need to estimate the covariance operator, which makes our methods computationally cheap. For example, our procedure can identify multiple change-points in O(nlog n) time. Our procedure is fully non-parametric and is robust to outliers through the use of data depth ranks. We show that when n is large, our methods have simple behaviour under the null hypothesis.We also show that the FKWC change-point procedures are n^-1/2-consistent. In addition to asymptotic results, we provide a finite sample accuracy result for our at-most-one change-point estimator. In simulation, we compare our methods against several others. We also present an application of our methods to intraday asset returns and f-MRI scans.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
11/18/2020

Robust, multiple change-point detection for covariance matrices using data depth

In this paper, two robust, nonparametric methods for multiple change-poi...
research
04/10/2023

Non-asymptotic inference for multivariate change point detection

Traditional methods for inference in change point detection often rely o...
research
05/05/2020

Homogeneity Tests of Covariance and Change-Points Identification for High-Dimensional Functional Data

We consider inference problems for high-dimensional (HD) functional data...
research
07/18/2020

Modeling Nonlinear Growth Followed by Long-Memory Equilibrium with Unknown Change Point

Measurements of many biological processes are characterized by an initia...
research
04/10/2018

Learning Latent Events from Network Message Logs: A Decomposition Based Approach

In this communication, we describe a novel technique for event mining us...
research
06/24/2020

Break Point Detection for Functional Covariance

Many experiments record sequential trajectories that oscillate around ze...
research
01/02/2018

ruptures: change point detection in Python

ruptures is a Python library for offline change point detection. This pa...

Please sign up or login with your details

Forgot password? Click here to reset