A mean-field stackelberg game approach for obfuscation adoption in empirical risk minimization

Jeffrey Pawlick, Quanyan Zhu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Data ecosystems are becoming larger and more complex, while privacy concerns are threatening to Erode their potential benefits. Recently, users have developed obfuscation techniques that issue fake search engine queries, undermine location tracking algorithms, or evade government surveillance. These techniques raise one conflict between each user and the machine learning algorithms which track the users, and one conflict between the users themselves. We use game theory to capture the first conflict with a Stackelberg game and the second conflict with a mean field game. Both are combined into a bi-level framework which quantifies accuracy using empirical risk minimization and privacy using differential privacy. We identify necessary and sufficient conditions under which 1) each user is incentivized to obfuscate if other users are obfuscating, 2) the tracking algorithm can avoid this by promising a level of privacy protection, and 3) this promise is incentive-compatible for the tracking algorithm.

Original languageEnglish (US)
Title of host publication2017 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages518-522
Number of pages5
Volume2018-January
ISBN (Electronic)9781509059904
DOIs
StatePublished - Mar 7 2018
Event5th IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017 - Montreal, Canada
Duration: Nov 14 2017Nov 16 2017

Other

Other5th IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017
CountryCanada
CityMontreal
Period11/14/1711/16/17

Fingerprint

Game theory
Search engines
Ecosystems
Learning algorithms
Learning systems

Keywords

  • Differential Privacy
  • Empirical Risk Minimization
  • Mean-Field Game
  • Obfuscation
  • Stackelberg Game

ASJC Scopus subject areas

  • Information Systems
  • Signal Processing

Cite this

Pawlick, J., & Zhu, Q. (2018). A mean-field stackelberg game approach for obfuscation adoption in empirical risk minimization. In 2017 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017 - Proceedings (Vol. 2018-January, pp. 518-522). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/GlobalSIP.2017.8308697

A mean-field stackelberg game approach for obfuscation adoption in empirical risk minimization. / Pawlick, Jeffrey; Zhu, Quanyan.

2017 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017 - Proceedings. Vol. 2018-January Institute of Electrical and Electronics Engineers Inc., 2018. p. 518-522.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Pawlick, J & Zhu, Q 2018, A mean-field stackelberg game approach for obfuscation adoption in empirical risk minimization. in 2017 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017 - Proceedings. vol. 2018-January, Institute of Electrical and Electronics Engineers Inc., pp. 518-522, 5th IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017, Montreal, Canada, 11/14/17. https://doi.org/10.1109/GlobalSIP.2017.8308697
Pawlick J, Zhu Q. A mean-field stackelberg game approach for obfuscation adoption in empirical risk minimization. In 2017 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017 - Proceedings. Vol. 2018-January. Institute of Electrical and Electronics Engineers Inc. 2018. p. 518-522 https://doi.org/10.1109/GlobalSIP.2017.8308697
Pawlick, Jeffrey ; Zhu, Quanyan. / A mean-field stackelberg game approach for obfuscation adoption in empirical risk minimization. 2017 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017 - Proceedings. Vol. 2018-January Institute of Electrical and Electronics Engineers Inc., 2018. pp. 518-522
@inproceedings{829cbf69f2644c03b7251d7846df7107,
title = "A mean-field stackelberg game approach for obfuscation adoption in empirical risk minimization",
abstract = "Data ecosystems are becoming larger and more complex, while privacy concerns are threatening to Erode their potential benefits. Recently, users have developed obfuscation techniques that issue fake search engine queries, undermine location tracking algorithms, or evade government surveillance. These techniques raise one conflict between each user and the machine learning algorithms which track the users, and one conflict between the users themselves. We use game theory to capture the first conflict with a Stackelberg game and the second conflict with a mean field game. Both are combined into a bi-level framework which quantifies accuracy using empirical risk minimization and privacy using differential privacy. We identify necessary and sufficient conditions under which 1) each user is incentivized to obfuscate if other users are obfuscating, 2) the tracking algorithm can avoid this by promising a level of privacy protection, and 3) this promise is incentive-compatible for the tracking algorithm.",
keywords = "Differential Privacy, Empirical Risk Minimization, Mean-Field Game, Obfuscation, Stackelberg Game",
author = "Jeffrey Pawlick and Quanyan Zhu",
year = "2018",
month = "3",
day = "7",
doi = "10.1109/GlobalSIP.2017.8308697",
language = "English (US)",
volume = "2018-January",
pages = "518--522",
booktitle = "2017 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017 - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - A mean-field stackelberg game approach for obfuscation adoption in empirical risk minimization

AU - Pawlick, Jeffrey

AU - Zhu, Quanyan

PY - 2018/3/7

Y1 - 2018/3/7

N2 - Data ecosystems are becoming larger and more complex, while privacy concerns are threatening to Erode their potential benefits. Recently, users have developed obfuscation techniques that issue fake search engine queries, undermine location tracking algorithms, or evade government surveillance. These techniques raise one conflict between each user and the machine learning algorithms which track the users, and one conflict between the users themselves. We use game theory to capture the first conflict with a Stackelberg game and the second conflict with a mean field game. Both are combined into a bi-level framework which quantifies accuracy using empirical risk minimization and privacy using differential privacy. We identify necessary and sufficient conditions under which 1) each user is incentivized to obfuscate if other users are obfuscating, 2) the tracking algorithm can avoid this by promising a level of privacy protection, and 3) this promise is incentive-compatible for the tracking algorithm.

AB - Data ecosystems are becoming larger and more complex, while privacy concerns are threatening to Erode their potential benefits. Recently, users have developed obfuscation techniques that issue fake search engine queries, undermine location tracking algorithms, or evade government surveillance. These techniques raise one conflict between each user and the machine learning algorithms which track the users, and one conflict between the users themselves. We use game theory to capture the first conflict with a Stackelberg game and the second conflict with a mean field game. Both are combined into a bi-level framework which quantifies accuracy using empirical risk minimization and privacy using differential privacy. We identify necessary and sufficient conditions under which 1) each user is incentivized to obfuscate if other users are obfuscating, 2) the tracking algorithm can avoid this by promising a level of privacy protection, and 3) this promise is incentive-compatible for the tracking algorithm.

KW - Differential Privacy

KW - Empirical Risk Minimization

KW - Mean-Field Game

KW - Obfuscation

KW - Stackelberg Game

UR - http://www.scopus.com/inward/record.url?scp=85048117923&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85048117923&partnerID=8YFLogxK

U2 - 10.1109/GlobalSIP.2017.8308697

DO - 10.1109/GlobalSIP.2017.8308697

M3 - Conference contribution

VL - 2018-January

SP - 518

EP - 522

BT - 2017 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2017 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

ER -