Paper 2023/068

Obfuscating Decision Trees

Shalini Banerjee, University of Auckland
Steven D. Galbraith, University of Auckland
Giovanni Russello, University of Auckland
Abstract

We construct a new encoder for hiding parameters in an interval membership function. As an interesting application, we design a simple and efficient virtual black-box obfuscator for evasive decision tree classifiers. The security of our construction relies upon random oracle paradigm. Our exclusive goal behind designing the obfuscator is that, not only will the solution increase the class of functions that has cryptographically secure obfuscators, but also address the open problem of non-interactive prediction in privacy-preserving classification using computationally inexpensive cryptographic hash functions.

Metadata
Available format(s)
PDF
Category
Cryptographic protocols
Publication info
Preprint.
Keywords
Program ObfuscationPrivacy-Preserving ClassificationDecision TreesVBB Security
Contact author(s)
shalini banerjee @ auckland ac nz
s galbraith @ auckland ac nz
g russello @ auckland ac nz
History
2023-02-09: last of 2 revisions
2023-01-20: received
See all versions
Short URL
https://ia.cr/2023/068
License
Creative Commons Attribution
CC BY

BibTeX

@misc{cryptoeprint:2023/068,
      author = {Shalini Banerjee and Steven D. Galbraith and Giovanni Russello},
      title = {Obfuscating Decision Trees},
      howpublished = {Cryptology ePrint Archive, Paper 2023/068},
      year = {2023},
      note = {\url{https://eprint.iacr.org/2023/068}},
      url = {https://eprint.iacr.org/2023/068}
}
Note: In order to protect the privacy of readers, eprint.iacr.org does not use cookies or embedded third party content.