Year: 2021
Type: article
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).
Authors: Jinhyun So, Basak Guler, A. Salman Avestimehr
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Awards: G1106929188, G1338515668, CIF: Medium: Collaborative Research: Coded Computing for Large-Scale Machine Learning, G4565307535, G4842797379, NETWORK SCIENCES: CODED COMPUTING: A TRANSFORMATIVE FRAMEWORK FOR TACTICAL WIRELESS EDGE COMPUTING, G799256164, CIF:Medium:Collaborative Research: Foundations of Coding for Modern Distributed Computing
OpenAlex ID: W2913183153
