Privacy-Preserving Contrastive Explanations with Local Foil Trees

Thijs Veugen; Bart Kamphorst; Michiel Marcus

doi:10.1007/978-3-031-07689-3_7

Privacy-Preserving Contrastive Explanations with Local Foil Trees

Thijs Veugen^*
, Bart Kamphorst
, Michiel Marcus

^*Corresponding author for this work

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

2 Citations (Scopus)

Abstract

We present the first algorithm that combines privacy-preserving technologies and state-of-the-art explainable AI to enable privacy-friendly explanations of black-box AI models. We provide a secure algorithm for contrastive explanations of black-box machine learning models that securely trains and uses local foil trees. Our work shows that the quality of these explanations can be upheld whilst ensuring the privacy of both the training data, and the model itself. An extended version of this paper is found at Cryptology ePrint Archive [16].

Original language	English
Title of host publication	Cyber Security, Cryptology, and Machine Learning
Subtitle of host publication	6th International Symposium, CSCML 2022, Be'er Sheva, Israel, June 30 – July 1, 2022, Proceedings
Editors	Shlomi Dolev, Amnon Meisels, Jonathan Katz
Place of Publication	Cham
Publisher	Springer
Pages	88-98
Number of pages	11
ISBN (Electronic)	978-3-031-07689-3
ISBN (Print)	978-3-031-07688-6
DOIs	https://doi.org/10.1007/978-3-031-07689-3_7
Publication status	Published - 2022
Externally published	Yes
Event	6th International Symposium on Cyber Security Cryptography and Machine Learning, CSCML 2022 - Beer Sheva, Israel Duration: 30 Jun 2022 → 1 Jul 2022 Conference number: 6

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
Volume	13301
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	6th International Symposium on Cyber Security Cryptography and Machine Learning, CSCML 2022
Abbreviated title	CSCML 2022
Country/Territory	Israel
City	Beer Sheva
Period	30/06/22 → 1/07/22

Keywords

Decision tree
Explainable AI
Foil tree
Secure multi-party computation
n/a OA procedure

Access to Document

10.1007/978-3-031-07689-3_7

Cite this

Veugen, T., Kamphorst, B., & Marcus, M. (2022). Privacy-Preserving Contrastive Explanations with Local Foil Trees. In S. Dolev, A. Meisels, & J. Katz (Eds.), Cyber Security, Cryptology, and Machine Learning: 6th International Symposium, CSCML 2022, Be'er Sheva, Israel, June 30 – July 1, 2022, Proceedings (pp. 88-98). (Lecture Notes in Computer Science; Vol. 13301). Springer. https://doi.org/10.1007/978-3-031-07689-3_7

Veugen, Thijs ; Kamphorst, Bart ; Marcus, Michiel. / Privacy-Preserving Contrastive Explanations with Local Foil Trees. Cyber Security, Cryptology, and Machine Learning: 6th International Symposium, CSCML 2022, Be'er Sheva, Israel, June 30 – July 1, 2022, Proceedings. editor / Shlomi Dolev ; Amnon Meisels ; Jonathan Katz. Cham : Springer, 2022. pp. 88-98 (Lecture Notes in Computer Science).

@inproceedings{bb8bd1c7fac440d78890b3ab6772fd55,

title = "Privacy-Preserving Contrastive Explanations with Local Foil Trees",

abstract = "We present the first algorithm that combines privacy-preserving technologies and state-of-the-art explainable AI to enable privacy-friendly explanations of black-box AI models. We provide a secure algorithm for contrastive explanations of black-box machine learning models that securely trains and uses local foil trees. Our work shows that the quality of these explanations can be upheld whilst ensuring the privacy of both the training data, and the model itself. An extended version of this paper is found at Cryptology ePrint Archive [16].",

keywords = "Decision tree, Explainable AI, Foil tree, Secure multi-party computation, n/a OA procedure",

author = "Thijs Veugen and Bart Kamphorst and Michiel Marcus",

note = "Funding Information: Acknowledgements. The research of this paper has been done within the FATE project, which is funded by the TNO Appl.AI program (internal AI program). We additionally thank Jasper van der Waa for his helpful comments and suggestions. Publisher Copyright: {\textcopyright} 2022, Springer Nature Switzerland AG.; 6th International Symposium on Cyber Security Cryptography and Machine Learning, CSCML 2022, CSCML 2022 ; Conference date: 30-06-2022 Through 01-07-2022",

year = "2022",

doi = "10.1007/978-3-031-07689-3\_7",

language = "English",

isbn = "978-3-031-07688-6",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "88--98",

editor = "Shlomi Dolev and Amnon Meisels and Jonathan Katz",

booktitle = "Cyber Security, Cryptology, and Machine Learning",

address = "Germany",

}

Veugen, T, Kamphorst, B & Marcus, M 2022, Privacy-Preserving Contrastive Explanations with Local Foil Trees. in S Dolev, A Meisels & J Katz (eds), Cyber Security, Cryptology, and Machine Learning: 6th International Symposium, CSCML 2022, Be'er Sheva, Israel, June 30 – July 1, 2022, Proceedings. Lecture Notes in Computer Science, vol. 13301, Springer, Cham, pp. 88-98, 6th International Symposium on Cyber Security Cryptography and Machine Learning, CSCML 2022, Beer Sheva, Israel, 30/06/22. https://doi.org/10.1007/978-3-031-07689-3_7

Privacy-Preserving Contrastive Explanations with Local Foil Trees. / Veugen, Thijs; Kamphorst, Bart; Marcus, Michiel.
Cyber Security, Cryptology, and Machine Learning: 6th International Symposium, CSCML 2022, Be'er Sheva, Israel, June 30 – July 1, 2022, Proceedings. ed. / Shlomi Dolev; Amnon Meisels; Jonathan Katz. Cham: Springer, 2022. p. 88-98 (Lecture Notes in Computer Science; Vol. 13301).

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

TY - GEN

T1 - Privacy-Preserving Contrastive Explanations with Local Foil Trees

AU - Veugen, Thijs

AU - Kamphorst, Bart

AU - Marcus, Michiel

N1 - Conference code: 6

PY - 2022

Y1 - 2022

N2 - We present the first algorithm that combines privacy-preserving technologies and state-of-the-art explainable AI to enable privacy-friendly explanations of black-box AI models. We provide a secure algorithm for contrastive explanations of black-box machine learning models that securely trains and uses local foil trees. Our work shows that the quality of these explanations can be upheld whilst ensuring the privacy of both the training data, and the model itself. An extended version of this paper is found at Cryptology ePrint Archive [16].

AB - We present the first algorithm that combines privacy-preserving technologies and state-of-the-art explainable AI to enable privacy-friendly explanations of black-box AI models. We provide a secure algorithm for contrastive explanations of black-box machine learning models that securely trains and uses local foil trees. Our work shows that the quality of these explanations can be upheld whilst ensuring the privacy of both the training data, and the model itself. An extended version of this paper is found at Cryptology ePrint Archive [16].

KW - Decision tree

KW - Explainable AI

KW - Foil tree

KW - Secure multi-party computation

KW - n/a OA procedure

UR - https://www.scopus.com/pages/publications/85134184139

U2 - 10.1007/978-3-031-07689-3_7

DO - 10.1007/978-3-031-07689-3_7

M3 - Conference contribution

AN - SCOPUS:85134184139

SN - 978-3-031-07688-6

T3 - Lecture Notes in Computer Science

SP - 88

EP - 98

BT - Cyber Security, Cryptology, and Machine Learning

A2 - Dolev, Shlomi

A2 - Meisels, Amnon

A2 - Katz, Jonathan

PB - Springer

CY - Cham

T2 - 6th International Symposium on Cyber Security Cryptography and Machine Learning, CSCML 2022

Y2 - 30 June 2022 through 1 July 2022

ER -

Veugen T, Kamphorst B, Marcus M. Privacy-Preserving Contrastive Explanations with Local Foil Trees. In Dolev S, Meisels A, Katz J, editors, Cyber Security, Cryptology, and Machine Learning: 6th International Symposium, CSCML 2022, Be'er Sheva, Israel, June 30 – July 1, 2022, Proceedings. Cham: Springer. 2022. p. 88-98. (Lecture Notes in Computer Science). doi: 10.1007/978-3-031-07689-3_7

Privacy-Preserving Contrastive Explanations with Local Foil Trees

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Privacy-Preserving Contrastive Explanations with Local Foil Trees

Cite this