摘要
数字化医疗健康数据的共享与分析对提升医疗服务质量和促进医疗科研至关重要,隐私保护成为其中关键挑战之一。由此,提出一种基于差分隐私的数字化医疗健康数据共享与分析隐私保护方法,通过在数据发布过程中添加噪声来保护个体隐私。在Adult数据集上与LDPKM(local differential privacy k-means)算法和DPLM(differential privacy local model)算法进行比较实验,验证了所提出的方法在保护隐私的同时,能够保持较高的数据分析准确性和有效性。
The sharing and analysis of digital health data are crucial for enhancing the quality of medical services and promoting medical research.Privacy preservation poses a significant challenge in this regard.Therefore,this paper proposes a privacy preservation mechanism for digital health data sharing and analysis based on differential privacy,which involves adding noise during the data publishing process to protect individual privacy.Comparative experiments with LDPKM(local differential privacy k-means)and DPLM(differential privacy local model)algorithms on the Adult dataset validate that the proposed method method can maintain high levels of data analysis accuracy and effectiveness while ensuring privacy preservation.
作者
马静
戴辉
MA Jing;DAI Hui(Information Center,Nanfang Hospital,Medical University Guangdong PR,Guangzhou 510515,China;Ganzhou Hospital-Nanfang Hospital,Southern Medical University(Ganzhou People’s Hospital),Ganzhou 341000,China)
出处
《微型电脑应用》
2025年第12期118-121,126,共5页
Microcomputer Applications
基金
赣州市科技计划项目(2023NS127395)。
关键词
差分隐私
数字化医疗健康
数据共享
隐私保护
differential privacy
digital health
data sharing
privacy preservation
