Segmentor3IsBack (version 1.5)

SelectModel: SelectModel

Description

This function is used to select the number of segments in the segmentation of the data according to several criterion including BIC, AIC, mBIC and oracle penalties.

Usage

SelectModel(x,penalty="oracle",seuil=n/log(n),keep=FALSE,greatjump=FALSE)

Arguments

x

An object of class Segmentor returned by function Segmentor

penalty

An object of class string stating which penalty to use in the model selection criteria. Accepted penalties are BIC, AIC, Zhang's modified BIC: mBIC and oracle penalties: 'oracle'. In the case of Normal segmentation, criterion developed by Lebarbier, in th

seuil

If penalty='oracle', an Integer for the threshold to use for the slope heuristic. Default value is n/log(n)

keep

A Boolean stating whether or not to keep the values of the criterion. Default is FALSE.

greatjump

If penalty is "oracle", a boolean stating whether to use the greatest jump (TRUE) or the threshold for the slope heuristic. Default is FALSE.

Value

• K
  The number of segments selected.
• crit
  If keep=TRUE, a vector of criterion value for each possible K.

Details

ll{ Package: Segmentor3IsBack Type: Package Version: 1.4 Date: 2013-03-25 License: GPL (>= 2) }

References

PDPA: Rigaill, G. Pruned dynamic programming for optimal multiple change-point detection: Submitted http://arxiv.org/abs/1004.0887

PDPA: Cleynen, A. and Koskas, M. and Rigaill, G. A Generic Implementation of the Pruned Dynamic Programing Algorithm: Submitted http://arxiv.org/abs/1204.5564

overdispersion parameter: Johnson, N. and Kemps, A. and Kotz, S. (2005) Univariate Discrete Distributions: John Wiley & Sons Selection criterion for counts: Cleynen, A. and Lebarbier, E. Segmentation of the Poisson and negative binomial rate models: a penalized estimator: Submitted http://arxiv.org/abs/1301.2534 Selection criterion for Gaussian distribution: Lebarbier, E. (2005) Detecting multiple change-points in the mean of Gaussian process by model selection: Signal Processing Slope heuristic: Arlot, S. and Bach, F. (2009) Data-driven calibration of penalties for least-squares regression: Journal of Machine Learning Research modified BIC: Zhang, N. and Siegmund, D. (2007) A modified Bayes information criterion with applications to the analysis of comparative genomic hybridization data: Biometrics

Examples

require(Segmentor3IsBack);
N=2000 
x=c(rnbinom(N,size=1.3,prob=0.7),rnbinom(N,size=1.3,prob=0.2),
rnbinom(N,size=1.3,prob=0.01), rnbinom(N,size=1.3,prob=0.2),
rnbinom(N,size=1.3,prob=0.8));
res=Segmentor(data=x,model=3,Kmax=20);  
# Finds the optimal segmentation in up to 20 segments with respect to 
#the negative binomial model.
Cr<-SelectModel(res,penalty='oracle',keep=FALSE)
Cr
#chooses the number of segments in the segmentation of x