Detecting relevant changes in time series models

DOI10.1111/RSSB.12121zbMATH Open1414.62360arXiv1403.8120OpenAlexW1923328694MaRDI QIDQ5378358FDOQ5378358

Holger Dette, Dominik Wied

Publication date: 12 June 2019

Published in: Journal of the Royal Statistical Society Series B: Statistical Methodology (Search for Journal in Brave)

Abstract: Most of the literature on change-point analysis by means of hypothesis testing considers hypotheses of the form H0 : heta_1 = heta_2 vs. H1 : heta_1 != heta_2, where heta_1 and heta_2 denote parameters of the process before and after a change point. This paper takes a different perspective and investigates the null hypotheses of no relevant changes, i.e. H0 : || heta_1 - heta_2|| ? leq Delta?, where || cdot || is an appropriate norm. This formulation of the testing problem is motivated by the fact that in many applications a modification of the statistical analysis might not be necessary, if the difference between the parameters before and after the change-point is small. A general approach to problems of this type is developed which is based on the CUSUM principle. For the asymptotic analysis weak convergence of the sequential empirical process has to be established under the alternative of non-stationarity, and it is shown that the resulting test statistic is asymptotically normal distributed. Several applications of the methodology are given including tests for relevant changes in the mean, variance, parameter in a linear regression model and distribution function among others. The finite sample properties of the new tests are investigated by means of a simulation study and illustrated by analyzing a data example from economics.

Full work available at URL: https://arxiv.org/abs/1403.8120

Recommendations

• Relevant change points in high dimensional time series
• Testing and estimating change-points in time series
• Tests for parameter changes in time series
• Detecting relevant changes in the mean of nonstationary processes -- a mass excess approach
• Consistent nonparametric change point detection combining CUSUM and marked empirical processes

zbMATH Keywords

strong mixingchange point analysisweak convergence under the alternativecumulative sumprecise hypothesesrelevant changes

Mathematics Subject Classification ID

Time series, auto-correlation, regression, etc. in statistics (GARCH) (62M10) Asymptotic distribution theory in statistics (62E20) Sequential statistical analysis (62L10) Non-Markovian processes: hypothesis testing (62M07)

Cited In (22)

• Monitoring mean changes in persistent multivariate time series
• Beyond Linear Dynamic Functional Connectivity: A Vine Copula Change Point Model
• Relevant change points in high dimensional time series
• Box-constrained monotone approximations to Lipschitz regularizations, with applications to robust testing
• Multiscale jump testing and estimation under complex temporal dynamics
• Detecting relevant changes in the mean of nonstationary processes -- a mass excess approach
• A computationally efficient nonparametric approach for changepoint detection
• Strong consistency for the conditional self-weighted \(M\) estimator of GRCA\((p)\) Models
• Functional data analysis in the Banach space of continuous functions
• Detecting relevant changes in the spatiotemporal mean function
• Testing for randomness in a random coefficient autoregression model
• Detecting structural breaks in eigensystems of functional time series
• A portmanteau-type test for detecting serial correlation in locally stationary functional time series
• Efficient change point detection and estimation in high-dimensional correlation matrices
• A Likelihood Ratio Approach to Sequential Change Point Detection for a General Class of Parameters
• Does modeling a structural break improve forecast accuracy?
• Change detection in autoregressive time series
• Mean stationarity test in time series: a signal variance-based approach
• Nonparametric Anomaly Detection on Time Series of Graphs
• Dynamic Semiparametric Factor Model With Structural Breaks
• Estimating a Change Point in a Sequence of Very High-Dimensional Covariance Matrices
• Time-based detection of changes to multivariate patterns

This page was built for publication: Detecting relevant changes in time series models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5378358)