标题： 面向高效的隐私保护机器学习：从协议、模型和系统视角的系统综述 显示英文标题

标题： Towards Efficient Privacy-Preserving Machine Learning: A Systematic Review from Protocol, Model, and System Perspectives

摘要： 基于密码学协议的隐私保护机器学习（PPML）已成为一种有前景的范式，以在基于云的机器学习服务中保护用户数据隐私。 虽然它实现了形式化的隐私保护，但由于与明文对应物相比存在数量级的开销，PPML通常会带来显著的效率和可扩展性成本。 因此，人们已经对减轻PPML的效率差距进行了大量研究。 在本次综述中，我们提供了一项全面且系统的回顾，重点是跨层次优化。 具体而言，我们将现有论文分为协议层、模型层和系统层，并回顾各层次的进展。 我们还根据技术见解对现有工作进行定性和定量比较，据此讨论未来的研究方向，并强调在协议、模型和系统层面上整合优化的必要性。 我们希望本次综述能够提供对现有方法的整体理解，并可能激发PPML领域的未来突破。 由于该领域发展迅速，我们还提供了一个公共的GitHub仓库，以持续跟踪发展，该仓库位于https://github.com/PKU-SEC-Lab/Awesome-PPML-Papers。 显示英文摘要

摘要： Privacy-preserving machine learning (PPML) based on cryptographic protocols has emerged as a promising paradigm to protect user data privacy in cloud-based machine learning services. While it achieves formal privacy protection, PPML often incurs significant efficiency and scalability costs due to orders of magnitude overhead compared to the plaintext counterpart. Therefore, there has been a considerable focus on mitigating the efficiency gap for PPML. In this survey, we provide a comprehensive and systematic review of recent PPML studies with a focus on cross-level optimizations. Specifically, we categorize existing papers into protocol level, model level, and system level, and review progress at each level. We also provide qualitative and quantitative comparisons of existing works with technical insights, based on which we discuss future research directions and highlight the necessity of integrating optimizations across protocol, model, and system levels. We hope this survey can provide an overarching understanding of existing approaches and potentially inspire future breakthroughs in the PPML field. As the field is evolving fast, we also provide a public GitHub repository to continuously track the developments, which is available at https://github.com/PKU-SEC-Lab/Awesome-PPML-Papers.