Privacy-Preserving Federated Unsupervised Domain Adaptation for Regression on Small-Scale and High-Dimensional Biological Data

2024-11-26

AI Summary

This work addresses the challenge of degraded generalization in regression tasks on small-sample, high-dimensional, cross-institutional biomedical data (e.g., DNA methylation), caused by domain shift and privacy constraints. We propose the first privacy-preserving federated unsupervised domain adaptation framework tailored for regression. Methodologically, we introduce Gaussian process regression into the federated learning setting for the first time, integrating randomized linear encoding with secure aggregation to enable cross-domain modeling without sharing raw data; unsupervised domain adaptation further mitigates distributional discrepancies across institutions. Evaluated on DNA methylation-based biological age prediction, our framework achieves performance comparable to centralized state-of-the-art methods while strictly ensuring data remain local—thereby establishing a novel paradigm for privacy-sensitive, distributed biomedical regression modeling.

Technology Category

• Machine Learning: Privacy
• Application Domains: Humanities & Computational Social Science
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Security and Privacy: Data transparency and provenance
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing

Abstract

Machine learning models often struggle with generalization in small, heterogeneous datasets due to domain shifts caused by variations in data collection and population differences. This challenge is particularly pronounced in biological data, where data is high-dimensional, small-scale, and decentralized across institutions. While federated domain adaptation methods (FDA) aim to address these challenges, most existing approaches rely on deep learning and focus on classification tasks, making them unsuitable for small-scale, high-dimensional applications. In this work, we propose freda, a privacy-preserving federated method for unsupervised domain adaptation in regression tasks. Unlike deep learning-based FDA approaches, freda is the first method to enable the federated training of Gaussian Processes to model complex feature relationships while ensuring complete data privacy through randomized encoding and secure aggregation. This allows for effective domain adaptation without direct access to raw data, making it well-suited for applications involving high-dimensional, heterogeneous datasets. We evaluate freda on the challenging task of age prediction from DNA methylation data, demonstrating that it achieves performance comparable to the centralized state-of-the-art method while preserving complete data privacy.

Problem

• Addresses domain shifts in biological data
• Enables privacy-preserving federated regression tasks
• Uses Gaussian Processes for complex feature relationships

Innovation

• Federated Gaussian Processes training
• Randomized encoding for privacy
• Secure aggregation for data protection