Nonconventional limit theorems in discrete and continuous time via martingales

Yuri Kifer, Srinivasa Varadhan

Research output: Contribution to journalArticle

Abstract

We obtain functional central limit theorems for both discrete time expressions of the form 1/√NΣ[Nt] n=1(F(X(q1(n)),., X(qℓ(n)))-F̄) and similar expressions in the continuous time where the sum is replaced by an integral. Here X(n),n≥0 is a sufficiently fast mixing vector process with some moment conditions and stationarity properties, F is a continuous function with polynomial growth and certain regularity properties, F̄=∫F d(μ×.×μ), μ is the distribution of X(0) and qi(n) = in for i≤k≤ℓ while for i > k they are positive functions taking on integer values on integers with some growth conditions which are satisfied, for instance, when qi 's are polynomials of increasing degrees. These results decisively generalize [Probab. Theory Related Fields 148 (2010) 71-106], whose method was only applicable to the case k = 2 under substantially more restrictive moment and mixing conditions and which could not be extended to convergence of processes and to the corresponding continuous time case. As in [Probab. Theory Related Fields 148 (2010) 71-106], our results hold true when Xi(n) = Tnfi, where T is a mixing subshift of finite type, a hyperbolic diffeomorphism or an expanding transformation taken with a Gibbs invariant measure, as well as in the case when Xi(n) = fi(γn), where γn is a Markov chain satisfying the Doeblin condition considered as a stationary process with respect to its invariant measure. Moreover, our relaxed mixing conditions yield applications to other types of dynamical systems and Markov processes, for instance, where a spectral gap can be established. The continuous time version holds true when, for instance, Xi(t) = fi(ξt), where ξt is a nondegenerate continuous time Markov chain with a finite state space or a nondegenerate diffusion on a compact manifold. A partial motivation for such limit theorems is due to a series of papers dealing with nonconventional ergodic averages.

Original languageEnglish (US)
Pages (from-to)649-688
Number of pages40
JournalAnnals of Probability
Volume42
Issue number2
DOIs
StatePublished - Mar 2014

Fingerprint

Limit Theorems
Martingale
Continuous Time
Mixing Conditions
Moment Conditions
Invariant Measure
Ergodic Averages
Functional Central Limit Theorem
Subshift
Integer
Continuous-time Markov Chain
Gibbs Measure
Polynomial Growth
Spectral Gap
Regularity Properties
Diffeomorphism
Stationarity
Finite Type
Stationary Process
Growth Conditions

Keywords

  • Hyperbolic diffeomorphisms
  • Limit theorems
  • Markov processes
  • Martingale approximation
  • Mixing

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Nonconventional limit theorems in discrete and continuous time via martingales. / Kifer, Yuri; Varadhan, Srinivasa.

In: Annals of Probability, Vol. 42, No. 2, 03.2014, p. 649-688.

Research output: Contribution to journalArticle

Kifer, Yuri ; Varadhan, Srinivasa. / Nonconventional limit theorems in discrete and continuous time via martingales. In: Annals of Probability. 2014 ; Vol. 42, No. 2. pp. 649-688.
@article{5126758d5f804ea09de62e1a0d757fe9,
title = "Nonconventional limit theorems in discrete and continuous time via martingales",
abstract = "We obtain functional central limit theorems for both discrete time expressions of the form 1/√NΣ[Nt] n=1(F(X(q1(n)),., X(qℓ(n)))-F̄) and similar expressions in the continuous time where the sum is replaced by an integral. Here X(n),n≥0 is a sufficiently fast mixing vector process with some moment conditions and stationarity properties, F is a continuous function with polynomial growth and certain regularity properties, F̄=∫F d(μ×.×μ), μ is the distribution of X(0) and qi(n) = in for i≤k≤ℓ while for i > k they are positive functions taking on integer values on integers with some growth conditions which are satisfied, for instance, when qi 's are polynomials of increasing degrees. These results decisively generalize [Probab. Theory Related Fields 148 (2010) 71-106], whose method was only applicable to the case k = 2 under substantially more restrictive moment and mixing conditions and which could not be extended to convergence of processes and to the corresponding continuous time case. As in [Probab. Theory Related Fields 148 (2010) 71-106], our results hold true when Xi(n) = Tnfi, where T is a mixing subshift of finite type, a hyperbolic diffeomorphism or an expanding transformation taken with a Gibbs invariant measure, as well as in the case when Xi(n) = fi(γn), where γn is a Markov chain satisfying the Doeblin condition considered as a stationary process with respect to its invariant measure. Moreover, our relaxed mixing conditions yield applications to other types of dynamical systems and Markov processes, for instance, where a spectral gap can be established. The continuous time version holds true when, for instance, Xi(t) = fi(ξt), where ξt is a nondegenerate continuous time Markov chain with a finite state space or a nondegenerate diffusion on a compact manifold. A partial motivation for such limit theorems is due to a series of papers dealing with nonconventional ergodic averages.",
keywords = "Hyperbolic diffeomorphisms, Limit theorems, Markov processes, Martingale approximation, Mixing",
author = "Yuri Kifer and Srinivasa Varadhan",
year = "2014",
month = "3",
doi = "10.1214/12-AOP796",
language = "English (US)",
volume = "42",
pages = "649--688",
journal = "Annals of Probability",
issn = "0091-1798",
publisher = "Institute of Mathematical Statistics",
number = "2",

}

TY - JOUR

T1 - Nonconventional limit theorems in discrete and continuous time via martingales

AU - Kifer, Yuri

AU - Varadhan, Srinivasa

PY - 2014/3

Y1 - 2014/3

N2 - We obtain functional central limit theorems for both discrete time expressions of the form 1/√NΣ[Nt] n=1(F(X(q1(n)),., X(qℓ(n)))-F̄) and similar expressions in the continuous time where the sum is replaced by an integral. Here X(n),n≥0 is a sufficiently fast mixing vector process with some moment conditions and stationarity properties, F is a continuous function with polynomial growth and certain regularity properties, F̄=∫F d(μ×.×μ), μ is the distribution of X(0) and qi(n) = in for i≤k≤ℓ while for i > k they are positive functions taking on integer values on integers with some growth conditions which are satisfied, for instance, when qi 's are polynomials of increasing degrees. These results decisively generalize [Probab. Theory Related Fields 148 (2010) 71-106], whose method was only applicable to the case k = 2 under substantially more restrictive moment and mixing conditions and which could not be extended to convergence of processes and to the corresponding continuous time case. As in [Probab. Theory Related Fields 148 (2010) 71-106], our results hold true when Xi(n) = Tnfi, where T is a mixing subshift of finite type, a hyperbolic diffeomorphism or an expanding transformation taken with a Gibbs invariant measure, as well as in the case when Xi(n) = fi(γn), where γn is a Markov chain satisfying the Doeblin condition considered as a stationary process with respect to its invariant measure. Moreover, our relaxed mixing conditions yield applications to other types of dynamical systems and Markov processes, for instance, where a spectral gap can be established. The continuous time version holds true when, for instance, Xi(t) = fi(ξt), where ξt is a nondegenerate continuous time Markov chain with a finite state space or a nondegenerate diffusion on a compact manifold. A partial motivation for such limit theorems is due to a series of papers dealing with nonconventional ergodic averages.

AB - We obtain functional central limit theorems for both discrete time expressions of the form 1/√NΣ[Nt] n=1(F(X(q1(n)),., X(qℓ(n)))-F̄) and similar expressions in the continuous time where the sum is replaced by an integral. Here X(n),n≥0 is a sufficiently fast mixing vector process with some moment conditions and stationarity properties, F is a continuous function with polynomial growth and certain regularity properties, F̄=∫F d(μ×.×μ), μ is the distribution of X(0) and qi(n) = in for i≤k≤ℓ while for i > k they are positive functions taking on integer values on integers with some growth conditions which are satisfied, for instance, when qi 's are polynomials of increasing degrees. These results decisively generalize [Probab. Theory Related Fields 148 (2010) 71-106], whose method was only applicable to the case k = 2 under substantially more restrictive moment and mixing conditions and which could not be extended to convergence of processes and to the corresponding continuous time case. As in [Probab. Theory Related Fields 148 (2010) 71-106], our results hold true when Xi(n) = Tnfi, where T is a mixing subshift of finite type, a hyperbolic diffeomorphism or an expanding transformation taken with a Gibbs invariant measure, as well as in the case when Xi(n) = fi(γn), where γn is a Markov chain satisfying the Doeblin condition considered as a stationary process with respect to its invariant measure. Moreover, our relaxed mixing conditions yield applications to other types of dynamical systems and Markov processes, for instance, where a spectral gap can be established. The continuous time version holds true when, for instance, Xi(t) = fi(ξt), where ξt is a nondegenerate continuous time Markov chain with a finite state space or a nondegenerate diffusion on a compact manifold. A partial motivation for such limit theorems is due to a series of papers dealing with nonconventional ergodic averages.

KW - Hyperbolic diffeomorphisms

KW - Limit theorems

KW - Markov processes

KW - Martingale approximation

KW - Mixing

UR - http://www.scopus.com/inward/record.url?scp=84894589507&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84894589507&partnerID=8YFLogxK

U2 - 10.1214/12-AOP796

DO - 10.1214/12-AOP796

M3 - Article

AN - SCOPUS:84894589507

VL - 42

SP - 649

EP - 688

JO - Annals of Probability

JF - Annals of Probability

SN - 0091-1798

IS - 2

ER -