On rate of convergence of statistical estimation of stationary ergodic processes

I. Csiszár, Zsolt Talata

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

Stationary ergodic processes with finite alphabets are approximated by finite memory processes based on an n-length realization of the process. Under the assumptions of summable continuity rate and non-nullness, a rate of convergence in d-distance is obtained, with explicit constants. Asymptotically, as n → ∞, the result is near the optimum.

Original languageEnglish
Article number5508630
Pages (from-to)3637-3641
Number of pages5
JournalIEEE Transactions on Information Theory
Volume56
Issue number8
DOIs
Publication statusPublished - Aug 2010

Fingerprint

Data storage equipment
continuity

Keywords

  • Finite memory estimators
  • infinite memory
  • Markov approximation
  • rate of convergence
  • stationary ergodic processes
  • statistical estimation

ASJC Scopus subject areas

  • Information Systems
  • Computer Science Applications
  • Library and Information Sciences

Cite this

On rate of convergence of statistical estimation of stationary ergodic processes. / Csiszár, I.; Talata, Zsolt.

In: IEEE Transactions on Information Theory, Vol. 56, No. 8, 5508630, 08.2010, p. 3637-3641.

Research output: Contribution to journalArticle

@article{54f59eee778d4a7abc2376511b44122e,
title = "On rate of convergence of statistical estimation of stationary ergodic processes",
abstract = "Stationary ergodic processes with finite alphabets are approximated by finite memory processes based on an n-length realization of the process. Under the assumptions of summable continuity rate and non-nullness, a rate of convergence in d-distance is obtained, with explicit constants. Asymptotically, as n → ∞, the result is near the optimum.",
keywords = "Finite memory estimators, infinite memory, Markov approximation, rate of convergence, stationary ergodic processes, statistical estimation",
author = "I. Csisz{\'a}r and Zsolt Talata",
year = "2010",
month = "8",
doi = "10.1109/TIT.2010.2050936",
language = "English",
volume = "56",
pages = "3637--3641",
journal = "IEEE Transactions on Information Theory",
issn = "0018-9448",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "8",

}

TY - JOUR

T1 - On rate of convergence of statistical estimation of stationary ergodic processes

AU - Csiszár, I.

AU - Talata, Zsolt

PY - 2010/8

Y1 - 2010/8

N2 - Stationary ergodic processes with finite alphabets are approximated by finite memory processes based on an n-length realization of the process. Under the assumptions of summable continuity rate and non-nullness, a rate of convergence in d-distance is obtained, with explicit constants. Asymptotically, as n → ∞, the result is near the optimum.

AB - Stationary ergodic processes with finite alphabets are approximated by finite memory processes based on an n-length realization of the process. Under the assumptions of summable continuity rate and non-nullness, a rate of convergence in d-distance is obtained, with explicit constants. Asymptotically, as n → ∞, the result is near the optimum.

KW - Finite memory estimators

KW - infinite memory

KW - Markov approximation

KW - rate of convergence

KW - stationary ergodic processes

KW - statistical estimation

UR - http://www.scopus.com/inward/record.url?scp=77954618804&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77954618804&partnerID=8YFLogxK

U2 - 10.1109/TIT.2010.2050936

DO - 10.1109/TIT.2010.2050936

M3 - Article

VL - 56

SP - 3637

EP - 3641

JO - IEEE Transactions on Information Theory

JF - IEEE Transactions on Information Theory

SN - 0018-9448

IS - 8

M1 - 5508630

ER -