As large-scale theft of data from corporate servers is becoming increasingly common, it becomes interesting to examine alternatives to the paradigm of centralizing sensitive data into large databases. Instead, one could use cryptography and distributed computation so that sensitive data can be supplied and processed in encrypted form, and only the final result is made known. In this paper, we examine how such a paradigm can be used to implement constraint satisfaction, a technique that can solve a broad class of AI problems such as resource allocation, planning, scheduling, and diagnosis. Most previous work on privacy in constraint satisfaction only attempted to protect specific types of information, in particular the feasibility of particular combinations of decisions. We formalize and extend these restricted notions of privacy by introducing four types of private information, including the feasibility of decisions and the final decisions made, but also the identities of the participants and the topology of the problem. We present distributed algorithms that allow computing solutions to constraint satisfaction problems while maintaining these four types of privacy. We formally prove the privacy properties of these algorithms, and show experiments that compare their respective performance on benchmark problems.
A Quantization-based Technique for Privacy Preserving Distributed Learning
Khalifa University（哈利法大学）；College of Engineering and IT, University of Dubai（迪拜大学工程与信息技术学院）；Univ. degli Studi di Milano（米兰大学）；EBTIC, Khalifa University（哈利法大学EBTIC）；C2PS, Khalifa University（哈利法大学C2PS）
Distributed Constraint Problems for Utilitarian Agents with Privacy Concerns, Recast as POMDPs
LAMIH UMR CNRS 8201, University of Valenciennes（瓦朗谢纳大学 LAMIH UMR CNRS 8201）；Kobe University（神户大学）；Kyushu University（九州大学）；Florida Institute of Technology（佛罗里达理工学院）
University of Valenciennes（瓦朗谢讷大学）；Nagoya Institute of Technology（名古屋工业大学）；Kobe University（神户大学）；Kyushu University（九州大学）；Florida Institute of Technology（佛罗里达理工学院）
Distributed Autonomous Online Learning: Regrets and Intrinsic Privacy-Preserving Properties
Purdue University（普渡大学）；University of Waterloo（滑铁卢大学）
在线学习在处理海量数据方面越来越受欢迎。然而，在线学习的顺序性质需要集中式学习器来存储数据和更新参数。在本文中，我们考虑在线学习与{ \em分布式}数据源。自主学习器根据本地数据源更新本地参数，并定期与通信网络中的一小部分邻居交换信息。我们推导了强凸函数的遗憾界，推广了Ram et al.（2010）关于凸函数的工作。最重要的是，我们表明，我们的算法具有\emph{intrinsic}隐私保护属性，我们证明了隐私保护的充分必要条件在网络中。这些条件意味着，对于具有大于一的连通性的网络，恶意学习者无法重建其他学习者的子梯度（和敏感的原始数据），这使得我们的算法在隐私敏感的应用程序中具有吸引力。
