• Corpus ID: 219401662

A Note on Online Change Point Detection

  title={A Note on Online Change Point Detection},
  author={Yi Yu and Oscar Hernan Madrid Padilla and Daren Wang and Alessandro Rinaldo},
  journal={arXiv: Statistics Theory},
Online change point detection is originated in sequential analysis, which has been thoroughly studied for more than half century. A variety of methods and optimality results have been established over the years. In this paper, we are concerned with the univariate online change point detection problem allowing for all model parameters to change. We establish a theoretical framework to allow for more refined minimax results. This includes a phase transition phenomenon and a study of the detection… 
Sequential change point detection in high dimensional time series
In a high dimensional scenario it is shown that the new monitoring schemes have asymptotic level alpha under the null hypothesis of no change and are consistent under the alternative of a change in at least one component.
A review on minimax rates in change point detection and localisation.
  • Yi Yu
  • Computer Science
  • 2020
This paper starts with the univariate mean change point analysis problem and review the state-of-the-art results in the literature, then moves on to more complex data types and investigates general principles behind the optimal procedures that lead to minimax rate-optimal results.
Optimal network online change point localisation
A minimax lower bound and two upper bounds based on NP-hard algorithms and polynomial-time algorithms, i.e. the detection delay, are established on the problem of online network change point detection.
Fast Online Changepoint Detection via Functional Pruning CUSUM statistics
An algorithm, Functional Online CuSUM (FO-CuS), which is equivalent to running these earlier methods simultaneously for all sizes of window, or all possible values for the size of change, and gives tight bounds on the expected computational cost per iteration of FOCuS.
Dynamic and heterogeneous treatment effects with abrupt changes
From personalised medicine to targeted advertising, it is an inherent task to provide a sequence of decisions with historical covariates and outcome data. This requires understanding of both the
Sequential Multivariate Change Detection with Calibrated and Memoryless False Detection Rates
This work presents a simulation-based approach to setting time-varying thresholds that allows a desired expected runtime to be accurately targeted whilst additionally keeping the false positive rate constant across time steps and shows how the cost of using the popular quadratic time MMD estimator can be reduced.
Locally private online change point detection
This work provides algorithms which respect the LDP constraint, which control the false alarm probability, and which detect changes with a minimal (minimax rate-optimal) delay, and presents the optimal rate in the benchmark, non-private setting.


Sequential change-point detection: Laplace concentration of scan statistics and non-asymptotic delay bounds
This work considers change-point detection in a fully sequential setup, and provides the first robust, finite-time analysis of the detection delay of this test that matches the known existing asymptotic orders, with fully explicit numerical constants.
Sequential changepoint detection in quality control and dynamical systems
A unified theory of sequential changepoint detection is introduced which leads to a class of sequential detection rules which are not too demanding in computational and memory requirements for on-line implementation and yet are nearly optimal under several performance criteria.
Delay time in sequential detection of change
High‐dimensional, multiscale online changepoint detection
A new method for high‐dimensional, online changepoint detection in settings where a p‐variate Gaussian data stream may undergo a change in mean, and it is proved that the patience, or average run length, of the procedure is at least at the desired nominal level.
A Likelihood Ratio Approach to Sequential Change Point Detection for a General Class of Parameters
It is proved that for a large class of testing problems the new detection scheme has asymptotic level α and is consistent and it is demonstrated that the new test performs better than the currently available procedures for these problems.
An online kernel change detection algorithm
This work considers the case where no accurate nor tractable model can be found, using a model-free approach, called Kernel change detection (KCD), and builds a dissimilarity measure in feature space between two sets of descriptors, shown to be asymptotically equivalent to the Fisher ratio in the Gaussian case.
A new approach for open‐end sequential change point monitoring
We propose a new sequential monitoring scheme for changes in the parameters of a multivariate time series. In contrast to procedures proposed in the literature which compare an estimator from the
Online change detection techniques in time series: An overview
Performance criteria including complexity, time granularity, and robustness is used to compare techniques, followed by a discussion about current challenges and open issues.
Information Bounds and Quick Detection of Parameter Changes in Stochastic Systems
  • T. Lai
  • Computer Science
    IEEE Trans. Inf. Theory
  • 1998
By using information-theoretic bounds and sequential hypothesis testing theory, this paper provides a new approach to optimal detection of abrupt changes in stochastic systems which leads to detection rules which have manageable computational complexity for on-line implementation and yet are nearly optimal under the different performance criteria considered.
Sequential change-point detection based on nearest neighbors
  • Hao Chen
  • Computer Science
    The Annals of Statistics
  • 2019
We propose a new framework for the detection of change-points in online, sequential data analysis. The approach utilizes nearest neighbor information and can be applied to sequences of multivariate