All Rivers Run to the Sea: Private Learning with Asymmetric Flows

arXiv:2312.05264v2 Announce Type: replace-cross
Abstract: Data privacy is of great concern in cloud machine-learning service platforms, when sensitive data are exposed to service providers. While private computing environments (e.g., secure enclaves), and cryptographic approaches (e.g., homomorphic encryption) provide strong privacy protection, their computing performance still falls short compared to cloud GPUs. To achieve privacy protection with high computing performance, we propose Delta, a new private training and inference framework, with comparable model performance as non-private centralized training. Delta features two …

abstract arxiv cloud computing cs.cr cs.lg data data privacy encryption environments homomorphic encryption machine performance platforms privacy protection rivers service service providers type