Strong Data Processing Constant Is Achieved by Binary Inputs

Or Ordentlich; Yury Polyanskiy

doi:10.1109/TIT.2021.3130189

Strong Data Processing Constant Is Achieved by Binary Inputs

Or Ordentlich
, Yury Polyanskiy

The Rachel and Selim Benin School of Engineering and Computer Science

The Hebrew University of Jerusalem

Research output: Contribution to journal › Article › peer-review

Abstract

For any channel P Y|X the strong data processing constant is defined as the smallest number η KL\in [0,1] such that I(U;Y)\≤ η KL I(U;X) holds for any Markov chain U-X-Y. It is shown that the value of η KL is given by that of the best binary-input subchannel of P Y|X. The same result holds for any f-divergence, verifying a conjecture of Cohen, Kemperman and Zbaganu (1998).

Original language	English
Pages (from-to)	1480-1481
Number of pages	2
Journal	IEEE Transactions on Information Theory
Volume	68
Issue number	3
DOIs	https://doi.org/10.1109/TIT.2021.3130189
State	Published - 1 Mar 2022

Keywords

Contraction coefficient
F-divergence
Strong data processing inequality (SDPI)

ASJC Scopus subject areas

Information Systems
Computer Science Applications
Library and Information Sciences

Access to Document

10.1109/TIT.2021.3130189

Cite this

@article{66d23b4284e547d59a24ab5838cc18c3,

title = "Strong Data Processing Constant Is Achieved by Binary Inputs",

abstract = "For any channel P Y|X the strong data processing constant is defined as the smallest number η KL\textbackslash{}in [0,1] such that I(U;Y)\textbackslash{}≤ η KL I(U;X) holds for any Markov chain U-X-Y. It is shown that the value of η KL is given by that of the best binary-input subchannel of P Y|X. The same result holds for any f-divergence, verifying a conjecture of Cohen, Kemperman and Zbaganu (1998).",

keywords = "Contraction coefficient, F-divergence, Strong data processing inequality (SDPI)",

author = "Or Ordentlich and Yury Polyanskiy",

note = "Publisher Copyright: {\textcopyright} 1963-2012 IEEE.",

year = "2022",

month = mar,

day = "1",

doi = "10.1109/TIT.2021.3130189",

language = "English",

volume = "68",

pages = "1480--1481",

journal = "IEEE Transactions on Information Theory",

issn = "0018-9448",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "3",

}

TY - JOUR

T1 - Strong Data Processing Constant Is Achieved by Binary Inputs

AU - Ordentlich, Or

AU - Polyanskiy, Yury

PY - 2022/3/1

Y1 - 2022/3/1

N2 - For any channel P Y|X the strong data processing constant is defined as the smallest number η KL\in [0,1] such that I(U;Y)\≤ η KL I(U;X) holds for any Markov chain U-X-Y. It is shown that the value of η KL is given by that of the best binary-input subchannel of P Y|X. The same result holds for any f-divergence, verifying a conjecture of Cohen, Kemperman and Zbaganu (1998).

AB - For any channel P Y|X the strong data processing constant is defined as the smallest number η KL\in [0,1] such that I(U;Y)\≤ η KL I(U;X) holds for any Markov chain U-X-Y. It is shown that the value of η KL is given by that of the best binary-input subchannel of P Y|X. The same result holds for any f-divergence, verifying a conjecture of Cohen, Kemperman and Zbaganu (1998).

KW - Contraction coefficient

KW - F-divergence

KW - Strong data processing inequality (SDPI)

UR - https://www.scopus.com/pages/publications/85120045737

U2 - 10.1109/TIT.2021.3130189

DO - 10.1109/TIT.2021.3130189

M3 - Article

SN - 0018-9448

VL - 68

SP - 1480

EP - 1481

JO - IEEE Transactions on Information Theory

JF - IEEE Transactions on Information Theory

IS - 3

ER -

Strong Data Processing Constant Is Achieved by Binary Inputs

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this