Advertisement
Published in:
2021 | OriginalPaper | Chapter
Public Verifiable Private Decision Tree Prediction
Authors : Hailong Wang, Yi Deng, Xiang Xie
Published in: Information Security and Cryptology
Publisher: Springer International Publishing
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
Abstract
Decision tree is a favored prediction model in machine learning and data mining. With the fast development and wide application of machine learning, the privacy of decision tree prediction is a rising concern.
In this paper, We construct a specific purpose NIZK for privacy-preserving decision tree prediction. The protocol allows the server who holds a decision tree model to convince others the result of the decision tree on an encrypted data sample, without leaking private information about the decision tree. Our protocol has high efficiency in both prover time and verifier time, and the proof size is only several KBs. With such NIZK, we can build a public verifiable private decision tree prediction system. In this system, a client can query the result of the server’s decision tree on its encrypted feature vector, and anyone who has only the access to public information can verify the validity of the result.