CodedPrivateML: A Fast and Privacy-Preserving Framework for Distributed Machine Learning
arXiv:1902.00641
Abstract
How to train a machine learning model while keeping the data private and secure? We present CodedPrivateML, a fast and scalable approach to this critical problem. CodedPrivateML keeps both the data and the model information-theoretically private, while allowing efficient parallelization of training across distributed workers. We characterize CodedPrivateML's privacy threshold and prove its convergence for logistic (and linear) regression. Furthermore, via extensive experiments on Amazon EC2, we demonstrate that CodedPrivateML provides significant speedup over cryptographic approaches based on multi-party computing (MPC).
Cited by in corpus (17)
- Lagrange Coded Computing: Optimal Design for Resiliency, Security and Privacy
- Minimizing Latency for Secure Coded Computing Using Secret Sharing via Staircase Codes
- Coded Machine Unlearning
- Turbo-Aggregate: Breaking the Quadratic Aggregation Barrier in Secure Federated Learning
- A Scalable Approach for Privacy-Preserving Collaborative Machine Learning
- Entangled Polynomial Codes for Secure, Private, and Batch Distributed Matrix Multiplication: Breaking the "Cubic" Barrier
- A Survey of Coded Distributed Computing
- VFL: A Verifiable Federated Learning with Privacy-Preserving for Big Data in Industrial IoT
- Privacy-Preserving Asynchronous Federated Learning Algorithms for Multi-Party Vertically Collaborative Learning
- DarKnight: A Data Privacy Scheme for Training and Inference of Deep Neural Networks
- Distributed Differentially Private Computation of Functions with Correlated Noise
- Improved Differentially Private Decentralized Source Separation for fMRI Data
- Edge Computing in the Dark: Leveraging Contextual-Combinatorial Bandit and Coded Computing
- Corella: A Private Multi Server Learning Approach based on Correlated Queries
- Privacy and Integrity Preserving Training Using Trusted Hardware
- Canoe : A System for Collaborative Learning for Neural Nets
- Coded Computing for Secure Boolean Computations