A Distributed Fair Machine Learning Framework with Private Demographic Data Protection

TL;DR

This paper introduces a distributed fair machine learning framework that protects private demographic data, enabling fair and accurate model training without exposing sensitive information, aligning with privacy regulations.

Contribution

It proposes a novel distributed framework for fair learning that preserves demographic privacy and demonstrates four effective methods with theoretical analysis.

Findings

Outperforms existing fair learning methods in fairness and accuracy

Successfully balances fairness and accuracy through a threshold parameter

Validated on three real-world datasets

Abstract

Fair machine learning has become a significant research topic with broad societal impact. However, most fair learning methods require direct access to personal demographic data, which is increasingly restricted to use for protecting user privacy (e.g. by the EU General Data Protection Regulation). In this paper, we propose a distributed fair learning framework for protecting the privacy of demographic data. We assume this data is privately held by a third party, which can communicate with the data center (responsible for model development) without revealing the demographic information. We propose a principled approach to design fair learning methods under this framework, exemplify four methods and show they consistently outperform their existing counterparts in both fairness and accuracy across three real-world data sets. We theoretically analyze the framework, and prove it can learn models with high fairness or high accuracy, with their trade-offs balanced by a threshold variable.