A Class of Nonbinary Symmetric Information Bottleneck Problems

Michael Dikshtein; Shlomo Shamai Shitz

doi:10.3929/ethz-b-000535270

A Class of Nonbinary Symmetric Information Bottleneck Problems

Michael Dikshtein, Shlomo Shamai Shitz

Technion - Israel Institute of Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

We study two dual settings of information pro cessing. Let Y → X → W be a Markov chain with fixed joint probability mass function PXY and a mutual information constraint on the pair (W, X). For the first problem, known as Information Bottleneck, we aim to maximize the mutual information between the random variables Y and W, while for the second problem, termed as Privacy Funnel, our goal is to minimize it. In particular, we analyze the scenario for which X is the input, and Y is the output of modulo-additive noise channel. We provide analytical characterization of the optimal information rates and the achieving distributions.

Original language	English
Title of host publication	International Zurich Seminar on Information and Communication (IZS 2022). Proceedings
Editors	Amos Lapidoth, Stefan M. Moser
Place of Publication	Zurich
Pages	9 - 13
DOIs	https://doi.org/10.3929/ethz-b-000535270
State	Published - 2022
Event	International Zurich Seminar on Information and Communication - Zurich Duration: 2 Mar 2022 → 4 Mar 2022 https://www.research-collection.ethz.ch/handle/20.500.11850/534535

Conference

Conference	International Zurich Seminar on Information and Communication
Abbreviated title	IZS
City	Zurich
Period	2/03/22 → 4/03/22
Internet address	https://www.research-collection.ethz.ch/handle/20.500.11850/534535

Access to Document

10.3929/ethz-b-000535270

Cite this

@inproceedings{2d231524bdef442a8da284bfe0de9ae1,

title = "A Class of Nonbinary Symmetric Information Bottleneck Problems",

abstract = "We study two dual settings of information pro cessing. Let Y → X → W be a Markov chain with fixed joint probability mass function PXY and a mutual information constraint on the pair (W, X). For the first problem, known as Information Bottleneck, we aim to maximize the mutual information between the random variables Y and W, while for the second problem, termed as Privacy Funnel, our goal is to minimize it. In particular, we analyze the scenario for which X is the input, and Y is the output of modulo-additive noise channel. We provide analytical characterization of the optimal information rates and the achieving distributions.",

author = "Michael Dikshtein and {Shamai Shitz}, Shlomo",

note = "International Zurich Seminar on Information and Communication (IZS 2022); Conference Location: Zurich, Switzerland; Conference Date: March 2–4, 2022; International Zurich Seminar on Information and Communication ; Conference date: 02-03-2022 Through 04-03-2022",

year = "2022",

doi = "10.3929/ethz-b-000535270",

language = "الإنجليزيّة",

pages = "9 -- 13",

editor = "Amos Lapidoth and Moser, {Stefan M.}",

booktitle = "International Zurich Seminar on Information and Communication (IZS 2022). Proceedings",

url = "https://www.research-collection.ethz.ch/handle/20.500.11850/534535",

}

TY - GEN

T1 - A Class of Nonbinary Symmetric Information Bottleneck Problems

AU - Dikshtein, Michael

AU - Shamai Shitz, Shlomo

N1 - International Zurich Seminar on Information and Communication (IZS 2022); Conference Location: Zurich, Switzerland; Conference Date: March 2–4, 2022

PY - 2022

Y1 - 2022

N2 - We study two dual settings of information pro cessing. Let Y → X → W be a Markov chain with fixed joint probability mass function PXY and a mutual information constraint on the pair (W, X). For the first problem, known as Information Bottleneck, we aim to maximize the mutual information between the random variables Y and W, while for the second problem, termed as Privacy Funnel, our goal is to minimize it. In particular, we analyze the scenario for which X is the input, and Y is the output of modulo-additive noise channel. We provide analytical characterization of the optimal information rates and the achieving distributions.

AB - We study two dual settings of information pro cessing. Let Y → X → W be a Markov chain with fixed joint probability mass function PXY and a mutual information constraint on the pair (W, X). For the first problem, known as Information Bottleneck, we aim to maximize the mutual information between the random variables Y and W, while for the second problem, termed as Privacy Funnel, our goal is to minimize it. In particular, we analyze the scenario for which X is the input, and Y is the output of modulo-additive noise channel. We provide analytical characterization of the optimal information rates and the achieving distributions.

U2 - 10.3929/ethz-b-000535270

DO - 10.3929/ethz-b-000535270

M3 - منشور من مؤتمر

SP - 9

EP - 13

BT - International Zurich Seminar on Information and Communication (IZS 2022). Proceedings

A2 - Lapidoth, Amos

A2 - Moser, Stefan M.

CY - Zurich

T2 - International Zurich Seminar on Information and Communication

Y2 - 2 March 2022 through 4 March 2022

ER -