Personalized and Private Peer-to-Peer Machine Learning

Abstract

The rise of connected personal devices together with privacy concerns call for machine learning algorithms capable of leveraging the data of a large number of agents to learn personalized models under strong privacy requirements. In this paper, we introduce an efficient algorithm to address the above problem in a fully decentralized (peer-to-peer) and asynchronous fashion, with provable convergence rate. We show how to make the algorithm differentially private to protect against the disclosure of information about the personal datasets, and formally analyze the trade-off between utility and privacy. Our experiments show that our approach dramatically outperforms previous work in the non-private case, and that under privacy constraints, we can significantly improve over models learned in isolation.

20 pages, to appear in the Proceedings of the 21st International Conference on Artificial Intelligence and Statistics (AISTATS 2018)

Cited by in corpus (12)

• Central Server Free Federated Learning over Single-sided Trust Social Networks
• The Internet of Federated Things (IoFT): A Vision for the Future and In-depth Survey of Data-driven Approaches for Federated Learning
• The OARF Benchmark Suite: Characterization and Implications for Federated Learning Systems
• FedCD: Improving Performance in non-IID Federated Learning
• Personalized Federated Learning for Intelligent IoT Applications: A Cloud-Edge based Framework
• Key Protected Classification for Collaborative Learning
• TEE-based decentralized recommender systems: The raw data sharing redemption
• A(DP)${}^2$SGD: Asynchronous Decentralized Parallel Stochastic Gradient Descent with Differential Privacy
• Trends and Advancements in Deep Neural Network Communication
• Muffliato: Peer-to-Peer Privacy Amplification for Decentralized Optimization and Averaging
• High-Dimensional Private Empirical Risk Minimization by Greedy Coordinate Descent
• Marketplace for AI Models