FDR-control in multiscale change-point segmentation

DOI10.1214/16-EJS1131zbMATH Open1338.62117arXiv1412.5844OpenAlexW3101925105WikidataQ57704067 ScholiaQ57704067MaRDI QIDQ153065FDOQ153065

Authors: Housen Li, Axel Munk, Hannes Sieling 

Publication date: 1 January 2016

Published in: Electronic Journal of Statistics (Search for Journal in Brave)

Abstract: Fast multiple change-point segmentation methods, which additionally provide faithful statistical statements on the number, locations and sizes of the segments, have recently received great attention. In this paper, we propose a multiscale segmentation method, FDRSeg, which controls the false discovery rate (FDR) in the sense that the number of false jumps is bounded linearly by the number of true jumps. In this way, it adapts the detection power to the number of true jumps. We prove a non-asymptotic upper bound for its FDR in a Gaussian setting, which allows to calibrate the only parameter of FDRSeg properly. Change-point locations, as well as the signal, are shown to be estimated in a uniform sense at optimal minimax convergence rates up to a log-factor. The latter is w.r.t. $L^p$-risk, $pge1$, over classes of step functions with bounded jump sizes and either bounded, or possibly increasing, number of change-points. FDRSeg can be efficiently computed by an accelerated dynamic program; its computational complexity is shown to be linear in the number of observations when there are many change-points. The performance of the proposed method is examined by comparisons with some state of the art methods on both simulated and real datasets. An R-package is available online.

Full work available at URL: https://arxiv.org/abs/1412.5844

Recommendations

• Multiscale change-point segmentation: beyond step functions
• Multiscale change-point segmentation: beyond step functions
• Segmentation and estimation of change-point models: false positive control and confidence regions
• On optimal segmentation and parameter tuning for multiple change-point detection and inference
• Change-point detection, segmentation, and related topics
• Multiscale change point detection
• Change-point testing for parallel data sets with FDR control
• FDR control of detected regions by multiscale matched filtering
• Wild binary segmentation for multiple change-point detection
• Fréchet change-point detection

zbMATH Keywords

false discovery ratedynamic programminghonest inferencearray CGH datachange-point regressiondeviation boundion channel recordingsminimax lower boundmultiscale inference

Mathematics Subject Classification ID

Nonparametric regression and quantile regression (62G08) Nonparametric hypothesis testing (62G10) Asymptotic properties of nonparametric inference (62G20) Dynamic programming (90C39)

Cites Work

• Local extremes, runs, strings and multiresolution. (With discussion)
• Wild binary segmentation for multiple change-point detection
• Weak convergence and empirical processes. With applications to statistics
• Pathwise coordinate optimization
• Ideal spatial adaptation by wavelet shrinkage
• Multiscale local change point detection with applications to value-at-risk
• Recursive computation of piecewise constant volatilities
• Title not available (Why is that?)
• Sparsity and Smoothness Via the Fused Lasso
• A Cluster Analysis Method for Grouping Means in the Analysis of Variance
• Consistencies and rates of convergence of jump-penalized least squares estimators
• Algorithms for the optimal identification of segment neighborhoods
• Multiscale testing of qualitative hypotheses
• The control of the false discovery rate in multiple testing under dependency.
• Multiple change-point detection via a screening and ranking algorithm
• Asymptotically optimal difference-based estimation of variance in nonparametric regression
• Title not available (Why is that?)
• Introduction to algorithms.
• Use of Cumulative Sums of Squares for Retrospective Detection of Changes of Variance
• Title not available (Why is that?)
• Multiple changepoint fitting via quasilikelihood, with application to DNA sequence segmentation
• Estimating and Testing Linear Models with Multiple Structural Changes
• Optimal detection of changepoints with a linear computational cost
• Detection with the scan and the average likelihood ratio
• Optimal sparse segment identification with application in copy number variation analysis
• Multiple Change-Point Estimation With a Total Variation Penalty
• Circular binary segmentation for the analysis of array-based DNA copy number data
• A Modified Bayes Information Criterion with Applications to the Analysis of Comparative Genomic Hybridization Data
• Introduction to nonparametric estimation
• The screening and ranking algorithm to detect DNA copy number variations
• Multiscale inference about a density
• False discovery rates and copy number variation
• Spatial smoothing and hot spot detection for CGH data using the fused lasso
• Bandwidth choice for nonparametric regression
• Tail probabilities for the null distribution of scanning statistics
• Extensions of smoothing via taut strings
• Minimal penalties for Gaussian model selection
• Optimal and fast detection of spatial clusters with scan statistics
• Strong invariance principles for dependent random variables
• Optimal detection of a jump in the intensity of a Poisson process or in a density with likelihood ratio statistics
• False discovery rate for scanning statistics
• Model selection for high dimensional multi-sequence change-point problems
• Title not available (Why is that?)
• Adaptive detection of multiple change-points in asset price volatility
• Estimating the Variance In Nonparametric Regression—What is a Reasonable Choice?
• Some comparisons for Gaussian processes
• Jump-Sparse and Sparse Recovery Using Potts Functionals
• Change-points: from sequential detection to biology and back
• Microarrays, empirical Bayes and the two-groups model

Cited In (30)

• Detecting multiple generalized change-points by isolating single ones
• Autocovariance estimation in regression with a discontinuous signal and \(m\)-dependent errors: a difference-based approach
• Change-point testing for parallel data sets with FDR control
• BayesProject: fast computation of a projection direction for multivariate changepoint detection
• Narrowest Significance Pursuit: Inference for Multiple Change-Points in Linear Models
• Multiscale change-point segmentation: beyond step functions
• Detection of multiple change-points in high-dimensional panel data with cross-sectional and temporal dependence
• Seeded intervals and noise level estimation in change point detection: a discussion of Fryzlewicz (2020)
• Multiple change point detection and validation in autoregressive time series data
• Activation discovery with FDR control: application to fMRI data
• Data-Driven Determination of the Number of Jumps in Regression Curves
• Cross-validation for change-point regression: pitfalls and solutions
• Multiscale Quantile Segmentation
• Multiscale change-point segmentation: beyond step functions
• Jump-penalized least absolute values estimation of scalar or circle-valued signals
• FDR control of detected regions by multiscale matched filtering
• On estimation of isotonic piecewise constant signals
• Moving Sum Data Segmentation for Stochastic Processes Based on Invariance
• Multiple testing of local extrema for detection of change points
• Bootstrap confidence intervals for multiple change points based on moving sum procedures
• Two-stage data segmentation permitting multiscale change points, heavy tails and dependence
• Narrowest-Over-Threshold Detection of Multiple Change Points and Change-Point-Like Features
• FDRSeg
• Bayesian sieve method for piece-wise smooth regression
• Detecting possibly frequent change-points: wild binary segmentation 2 and steepest-drop model selection
• Multidimensional multiscale scanning in exponential families: limit theory and statistical consequences
• Data-driven selection of the number of change-points via error rate control
• Relating and comparing methods for detecting changes in mean
• Consistency of a range of penalised cost approaches for detecting multiple changepoints
• Tail-greedy bottom-up data decompositions and fast multiple change-point detection

Uses Software

• R
• wbs
• ftnonpar
• FDRSeg
• dSTEM

This page was built for publication: FDR-control in multiscale change-point segmentation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q153065)