How to Democratise and Protect AI: Fair and Differentially Private Decentralised Deep Learning

TL;DR

This paper introduces FDPDDL, a decentralized deep learning framework that combines fairness and differential privacy through a reputation system, DPGAN, and DPSGD, enabling accurate, fair, and privacy-preserving collaborative model training.

Contribution

It proposes a novel two-stage scheme integrating reputation, fairness, and differential privacy in decentralized deep learning, which is a significant advancement over existing methods.

Findings

FDPDDL achieves high fairness in collaborative learning.

The framework yields accuracy comparable to centralized methods.

It outperforms standalone models in accuracy while preserving privacy.

Abstract

This paper firstly considers the research problem of fairness in collaborative deep learning, while ensuring privacy. A novel reputation system is proposed through digital tokens and local credibility to ensure fairness, in combination with differential privacy to guarantee privacy. In particular, we build a fair and differentially private decentralised deep learning framework called FDPDDL, which enables parties to derive more accurate local models in a fair and private manner by using our developed two-stage scheme: during the initialisation stage, artificial samples generated by Differentially Private Generative Adversarial Network (DPGAN) are used to mutually benchmark the local credibility of each party and generate initial tokens; during the update stage, Differentially Private SGD (DPSGD) is used to facilitate collaborative privacy-preserving deep learning, and local credibility and tokens of each party are updated according to the quality and quantity of individually released gradients. Experimental results on benchmark datasets under three realistic settings demonstrate that FDPDDL achieves high fairness, yields comparable accuracy to the centralised and distributed frameworks, and delivers better accuracy than the standalone framework.