Privacy Preserving Collaborative Deep Learning Using Verifiable Multi-Secret Sharing Scheme
Wulan Sri Lestari(a*) Ronsen Purba (a) Arwin Halim (a)
a) Magister Teknologi Informasi Department, STMIK Mikroskil
Jalan Thamrin 112, Medan 20212, Indonesia
wulan.lestari[at]mikroskil.ac.id
Abstract
Collaborative deep learning is an approach that used to overcome the amount of training data needed in building a better deep learning model. In collaborative deep learning, the central server collects user data and run the deep learning algorithm centrally to get more accurate models. However, centralized training data collection can raise serious of privacy leakage problem and damage to the integrity of training data. In this paper we introduce the privacy preserving collaborative deep learning model using verifiable (k, t, n) multi-secret sharing based on the Elliptic Curve Diffie Helman and SHA3-256 as a hash function. Where all training data will be formed into n shares using a session key generated from the private key and public key Elliptic Curve Diffie helman to protect the privacy and avoid all training data using SHA3-256 for the verification process before sending to server. The test results show the integrity of damaged training data and colluding participants can be verified using the Elliptic Curve Diffie Helman and SHA3-256. Therefore proposed model can protect the privacy and integrity of training data and maintain the accuracy of the deep learning model.
Keywords: Data Privacy; Data Integrity; Collaborative Deep Learning; Verifiable Multi-Secret Sharing
Topic: Computer Science and Engineering