China
Africa
Explore chapters and articles related to this topic
Differential privacy for edge computing-based smart grid operating over blockchain
Published in Muhammad Maaz Rehan, Mubashir Husain Rehmani, Blockchain-enabled Fog and Edge Computing, 2020
Muneeb Ul Hassan, Mubashir Husain Rehmani, Jinjun Chen
The second function is the algorithm, On-Demand Reporting; this function is the core part of the BDOR strategy because of its dynamic differentially private nature. In this function, the grid utility is given the privilege of asking queries, such as ‘How many houses are generating energy more than 500 w at a specific time slot?’ or ‘What is the maximum amount of energy being generated by a specific house right now?’ and many other similar queries. These types of queries are highly private, and if the utility becomes an adversary, it can leak critical private information. Therefore, we use exponential mechanism of differential privacy during this query evaluation to protect users’ privacy. In this reporting function, we first allowed the utility to ask queries, and when one utility ask queries from the aggregator database, then all required data related to a specific query is gathered autonomously. Afterward, an exponential differentially private functions makes a differentially private distribution for all input values within a range of a specific private parameter and a sensitivity value ∆q. The formula used to formulise the distribution is as follows: () Pr(F(q,S,UID)=UID)∝exp(ε.q(S,UID)2Δq)∑UID′∈S(UID)exp(ε.q(S,U′ID)2Δq)
DPWeVote: differentially private weighted voting protocol for cloud-based decision-making
Published in Enterprise Information Systems, 2019
Ziqi Yan, Jiqiang Liu, Shaowu Liu
From the aspects of the recent popular privacy model, differential privacy, McSherry and Talwar (2007) initially built the bridge between mechanism design and differential privacy, and proposed the Exponential mechanism especially for private auction. There are some subsequent researches on private auction and its truthfulness such as (Nissim, Smorodinsky, and Tennenholtz 2012; Xiao 2013). For private voting, in (Chen et al. 2013), they proposed a novel way to incorporate differential privacy directly into the player’s utility functions, a private two-candidate elections mechanism was designed. Leung and Lui (2012) considered the Bayesian setting of the distribution of the players types and proposed the Bayesian differential privacy while achieving persistent approximate truthfulness. Lee (2015) proposed an algorithm that satisfies both -differential privacy and -strategyproof for protecting participant privacy in tournament voting rules. Recently, considering the rank aggregation scenario in which the data curator is trusted, Hay, Elagina, and Miklau (2017) extended three non-private rank aggregation algorithms to their differentially private versions.
Deep Learning: Differential Privacy Preservation in the Era of Big Data
Published in Journal of Computer Information Systems, 2023
The exponential mechanism provides randomized results for non-numeric queries merged with the score function to examine the quality of the output S. Assume that is the score function of the dataset , which evaluates the output quality and is the sensitivity of . The exponential mechanism satisfies -differential privacy by the following equation.