Estimation of the Variance of Partial Sums of Dependent Processes

Daniel Vogel, Herold Dehling, Roland Fried, Olimjon Sharipov, Max Wornowizki

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

We study subsampling estimators for the limit variance σ2=V ar(X1)+2∑∞k=2Cov(X1,Xk) of partial sums of a stationary stochastic process (Xk)k≥1. We establish L2-consistency of a non-overlapping block resampling method. Our results apply to processes that can be represented as functionals of strongly mixing processes. Motivated by recent applications to rank tests, we also study estimators for the series V ar(F(X1))+2∑∞k=2Cov(F(X1),F(Xk)), where F is the distribution function of X1. Simulations illustrate the usefulness of the proposed estimators and of a mean squared error optimal rule for the choice of the block length.
Original languageEnglish
Pages (from-to)141-147
Number of pages7
JournalStatistics and Probability Letters
Volume83
Issue number1
Early online date23 Aug 2012
DOIs
Publication statusPublished - Jan 2013

Fingerprint

Partial Sums
Estimator
Dependent
Strongly Mixing
Resampling Methods
Subsampling
Rank Test
Mixing Processes
Block Method
Stationary Process
Mean Squared Error
Stochastic Processes
Distribution Function
Series
Simulation

Keywords

  • Variance extimation
  • Weakly dependent Processes
  • Subsampling techniques
  • Central limit theorem
  • Functionals of mixing processes

Cite this

Estimation of the Variance of Partial Sums of Dependent Processes. / Vogel, Daniel; Dehling, Herold; Fried, Roland ; Sharipov, Olimjon; Wornowizki, Max.

In: Statistics and Probability Letters, Vol. 83, No. 1, 01.2013, p. 141-147.

Research output: Contribution to journalArticle

Vogel, Daniel ; Dehling, Herold ; Fried, Roland ; Sharipov, Olimjon ; Wornowizki, Max. / Estimation of the Variance of Partial Sums of Dependent Processes. In: Statistics and Probability Letters. 2013 ; Vol. 83, No. 1. pp. 141-147.
@article{3108ead851a040fd81482d41b19766c2,
title = "Estimation of the Variance of Partial Sums of Dependent Processes",
abstract = "We study subsampling estimators for the limit variance σ2=V ar(X1)+2∑∞k=2Cov(X1,Xk) of partial sums of a stationary stochastic process (Xk)k≥1. We establish L2-consistency of a non-overlapping block resampling method. Our results apply to processes that can be represented as functionals of strongly mixing processes. Motivated by recent applications to rank tests, we also study estimators for the series V ar(F(X1))+2∑∞k=2Cov(F(X1),F(Xk)), where F is the distribution function of X1. Simulations illustrate the usefulness of the proposed estimators and of a mean squared error optimal rule for the choice of the block length.",
keywords = "Variance extimation, Weakly dependent Processes, Subsampling techniques, Central limit theorem, Functionals of mixing processes",
author = "Daniel Vogel and Herold Dehling and Roland Fried and Olimjon Sharipov and Max Wornowizki",
note = "The authors were supported in part by the Collaborative Research Grant 823, Project C3 Analysis of Structural Change in Dynamic Processes, of the German Research Foundation.",
year = "2013",
month = "1",
doi = "10.1016/j.spl.2012.08.012",
language = "English",
volume = "83",
pages = "141--147",
journal = "Statistics and Probability Letters",
issn = "0167-7152",
publisher = "Elsevier",
number = "1",

}

TY - JOUR

T1 - Estimation of the Variance of Partial Sums of Dependent Processes

AU - Vogel, Daniel

AU - Dehling, Herold

AU - Fried, Roland

AU - Sharipov, Olimjon

AU - Wornowizki, Max

N1 - The authors were supported in part by the Collaborative Research Grant 823, Project C3 Analysis of Structural Change in Dynamic Processes, of the German Research Foundation.

PY - 2013/1

Y1 - 2013/1

N2 - We study subsampling estimators for the limit variance σ2=V ar(X1)+2∑∞k=2Cov(X1,Xk) of partial sums of a stationary stochastic process (Xk)k≥1. We establish L2-consistency of a non-overlapping block resampling method. Our results apply to processes that can be represented as functionals of strongly mixing processes. Motivated by recent applications to rank tests, we also study estimators for the series V ar(F(X1))+2∑∞k=2Cov(F(X1),F(Xk)), where F is the distribution function of X1. Simulations illustrate the usefulness of the proposed estimators and of a mean squared error optimal rule for the choice of the block length.

AB - We study subsampling estimators for the limit variance σ2=V ar(X1)+2∑∞k=2Cov(X1,Xk) of partial sums of a stationary stochastic process (Xk)k≥1. We establish L2-consistency of a non-overlapping block resampling method. Our results apply to processes that can be represented as functionals of strongly mixing processes. Motivated by recent applications to rank tests, we also study estimators for the series V ar(F(X1))+2∑∞k=2Cov(F(X1),F(Xk)), where F is the distribution function of X1. Simulations illustrate the usefulness of the proposed estimators and of a mean squared error optimal rule for the choice of the block length.

KW - Variance extimation

KW - Weakly dependent Processes

KW - Subsampling techniques

KW - Central limit theorem

KW - Functionals of mixing processes

U2 - 10.1016/j.spl.2012.08.012

DO - 10.1016/j.spl.2012.08.012

M3 - Article

VL - 83

SP - 141

EP - 147

JO - Statistics and Probability Letters

JF - Statistics and Probability Letters

SN - 0167-7152

IS - 1

ER -