Finding multiple abrupt change points

Venter, J. H.; Steel, S. J.

doi:10.1016/0167-9473(96)00007-2

Cited by 28 publications

(14 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From this table, we may also calculate any of several measures of the degree to which ones and zeros are distributed randomly in the segments defined by the given segmentation. For example, we may determine values for chi-square (equivalent to the mean-squared error measure used by Venter and Steel (1996), as well as an extension of the Anderson-Darling statistic to multiple changepoints) , Fisher's exact test, likelihood ratios under the null hypothesis and under the alternative hypothesis indicated by the given segmentation, or a generalization of the Kolmogorov-Smirnov two-sample statistic. These measures have been employed previously for the problem of detecting a single changepoint in a binary sequence (see Halpern, 1999 for discussion of differences between measures).…”

Section: Testing For Multiple Changepointsmentioning

confidence: 99%

“…As Finally, our third question: Should we reject Ho? As discussed by Venter and Steel (1996), the Bonferroni argument gives (Nl)PN as a conservative estimate of the probability of obtaining at least as small a value of PN as that of our sequence of interest by chance. For the current implementation, concerned with binary sequences with an alternating pattern of segment types, the bound might be acceptable for N less than four or five but becomes excessively conservative as N rises.…”

Section: Testing For Multiple Changepointsmentioning

confidence: 99%

“…The number of segments is not known in advance. Venter and Steel (1996) provided simulation-based methods for solving a related problem, that of a sequence of continuously valued observations that may be composed of some unknown number of segments, each of which has its own mean. It is shown here that related methods provide a solution for the alternating segments problem for binary sequences.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Multiple‐Changepoint Testing for an Alternating Segments Model of a Binary Sequence

Halpern

2000

Biometrics

View full text Add to dashboard Cite

A binary sequence may give the appearance of being composed of alternating segments with relatively high and relatively low probability of success. Determining whether such an alternating pattern is significant is a multiple-changepoint problem where the number of segments and their success probabilities are unknown, with the added constraint of segment alternation. A dynamic programming method for determining the optimal segmentation into a given number of segments is provided. Given this, a variation on the simulation method of Venter and Steel (1996, Computational Statistics and Data Analysis 22, 481-504) may be employed to test the null hypothesis of a homogeneous sequence as well as to estimate the number and location of changepoints. A sample application, the assessment of the possibility of genetic recombination in HIV sequences, is presented.

show abstract

Section: Testing For Multiple Changepointsmentioning

confidence: 99%

Section: Testing For Multiple Changepointsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Multiple‐Changepoint Testing for an Alternating Segments Model of a Binary Sequence

Halpern

2000

Biometrics

View full text Add to dashboard Cite

show abstract

“…There is a substantial statistical literature on methods for detecting change points. Many early methods relied on classic significance testing to find changes, using standard parametric and non-parametric tests (e.g., Pettitt, 1979;Venter & Steel, 1996). Bayesian approaches were also developed relatively early.…”

Section: Introductionmentioning

confidence: 99%

A simple and flexible Bayesian method for inferring step changes in cognition

Lee

2018

Behav Res

View full text Add to dashboard Cite

Human behavioral data often shows patterns of sudden change over time. Sometimes the causes of these step changes are internal, such as learning curves changing abruptly when a learner implements a new rule. Sometimes the cause is external, such a people's opinions about a topic changing in response to a new relevant event. Detecting change points in sequences of binary data is a basic statistical problem, with many existing solutions, but they seem rarely to be used in psychological modeling. We develop a simple and flexible Bayesian approach to modeling step changes in cognition, implemented as a graphical model in JAGS. The model is able to infer how many change points are justified by the data, as well as the location of the change points. The basic model is also easily extended to include latent-mixture and hierarchical structures, allowing it to be tailored to specific cognitive modeling problems. We demonstrate the adequacy of the basic model by applying it to the classic Lindisfarne Scribes problem, and the flexibility of the modeling approach through two new applications. The first involves a latent-mixture model to determine if individuals learn categories incrementally or in discrete stages. The second involves a hierarchical model of crowd-sourced predictions about the winner of the U.S. National Football League's Most Valuable Player for the 2016-2017 season.

show abstract

“…Belisle et al (1998) made inferences about the Bayesian hierarchical change-point model with an ensemble of sample paths for neuron spike train data. In the multiple change-point setting, Venter and Steel (1996) identified multiple abrupt change-points in a sequence of observations via hypothesis testing. Hawkins (2001) developed an approach with maximum likelihood estimates of the change-points and within-segment parameters in the exponential family.…”

Section: Introductionmentioning

confidence: 99%

Bayesian Multiple Change-Point Estimation and Segmentation

Kim¹,

Cheon²

2013

Communications for Statistical Applications and Methods

View full text Add to dashboard Cite

This study presents a Bayesian multiple change-point detection approach to segment and classify the observations that no longer come from an initial population after a certain time. Inferences are based on the multiple change-points in a sequence of random variables where the probability distribution changes. Bayesian multiple change-point estimation is classifies each observation into a segment. We use a truncated Poisson distribution for the number of change-points and conjugate prior for the exponential family distributions. The Bayesian method can lead the unsupervised classification of discrete, continuous variables and multivariate vectors based on latent class models; therefore, the solution for change-points corresponds to the stochastic partitions of observed data. We demonstrate segmentation with real data.

show abstract

Finding multiple abrupt change points

Cited by 28 publications

References 12 publications

Multiple‐Changepoint Testing for an Alternating Segments Model of a Binary Sequence

Multiple‐Changepoint Testing for an Alternating Segments Model of a Binary Sequence

A simple and flexible Bayesian method for inferring step changes in cognition

Bayesian Multiple Change-Point Estimation and Segmentation

Contact Info

Product

Resources

About