Generating Synthetic Health Sensor Data for Privacy-Preserving Wearable Stress Detection

被引:0
|
作者
Lange, Lucas [1 ]
Wenzlitschke, Nils [1 ]
Rahm, Erhard [1 ]
机构
[1] Univ Leipzig, ScaDS AI Dresden Leipzig, Augustuspl 10, D-04109 Leipzig, Germany
关键词
generative adversarial network; stress recognition; privacy-preserving machine learning; differential privacy; smartwatch; time series; physiological sensor data; synthetic data; smart health;
D O I
10.3390/s24103052
中图分类号
O65 [分析化学];
学科分类号
070302 ; 081704 ;
摘要
Smartwatch health sensor data are increasingly utilized in smart health applications and patient monitoring, including stress detection. However, such medical data often comprise sensitive personal information and are resource-intensive to acquire for research purposes. In response to this challenge, we introduce the privacy-aware synthetization of multi-sensor smartwatch health readings related to moments of stress, employing Generative Adversarial Networks (GANs) and Differential Privacy (DP) safeguards. Our method not only protects patient information but also enhances data availability for research. To ensure its usefulness, we test synthetic data from multiple GANs and employ different data enhancement strategies on an actual stress detection task. Our GAN-based augmentation methods demonstrate significant improvements in model performance, with private DP training scenarios observing an 11.90-15.48% increase in F1-score, while non-private training scenarios still see a 0.45% boost. These results underline the potential of differentially private synthetic data in optimizing utility-privacy trade-offs, especially with the limited availability of real training samples. Through rigorous quality assessments, we confirm the integrity and plausibility of our synthetic data, which, however, are significantly impacted when increasing privacy requirements.
引用
收藏
页数:24
相关论文
共 50 条
  • [1] Privacy-Preserving Anomaly Detection Using Synthetic Data
    Mayer, Rudolf
    Hittmeir, Markus
    Ekelhart, Andreas
    DATA AND APPLICATIONS SECURITY AND PRIVACY XXXIV, DBSEC 2020, 2020, 12122 : 195 - 207
  • [2] A Secure Privacy-Preserving Data Aggregation Model in Wearable Wireless Sensor Networks
    Zhang, Changlun
    Li, Chao
    Zhang, Jian
    JOURNAL OF ELECTRICAL AND COMPUTER ENGINEERING, 2015, 2015
  • [3] Generating Privacy Preserving Synthetic Medical Data
    Faisal, Fahim
    Mohammed, Noman
    Leung, Carson K.
    Wang, Yang
    2022 IEEE 9TH INTERNATIONAL CONFERENCE ON DATA SCIENCE AND ADVANCED ANALYTICS (DSAA), 2022, : 1003 - 1012
  • [4] A systematic review of privacy-preserving techniques for synthetic tabular health data
    Tobias Hyrup
    Anton D. Lautrup
    Arthur Zimek
    Peter Schneider-Kamp
    Discover Data, 3 (1):
  • [5] Privacy-Preserving Synthetic Educational Data Generation
    Vie, Jill-Jenn
    Rigaux, Tomas
    Minn, Sein
    EDUCATING FOR A NEW FUTURE: MAKING SENSE OF TECHNOLOGY-ENHANCED LEARNING ADOPTION, EC-TEL 2022, 2022, 13450 : 393 - 406
  • [6] Privacy-Preserving Synthetic Smart Meters Data
    Del Grosso, Ganesh
    Pichler, Georg
    Piantanida, Pablo
    2021 IEEE POWER & ENERGY SOCIETY INNOVATIVE SMART GRID TECHNOLOGIES CONFERENCE (ISGT), 2021,
  • [7] Generating bimodal privacy-preserving data for face recognition
    Tomasevic, Darian
    Boutros, Fadi
    Damer, Naser
    Peer, Peter
    Struc, Vitomir
    ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE, 2024, 133
  • [8] Development of privacy-preserving sensor for person detection
    Nakashima, Shota
    Kitazono, Yuhki
    Zhang, Lifeng
    Serikawa, Seiichi
    1ST INTERNATIONAL CONFERENCE ON SECURITY CAMERA NETWORK, PRIVACY PROTECTION AND COMMUNITY SAFETY 2009, 2010, 2 (01): : 213 - 217
  • [9] Privacy-Preserving Sharing of Mobile Sensor Data
    Liu, Yin
    Cruz, Breno Dantas
    Tilevich, Eli
    MOBILE COMPUTING, APPLICATIONS, AND SERVICES, MOBICASE 2021, 2022, 434 : 19 - 41
  • [10] Trusted and privacy-preserving sensor data onloading
    Liu, Yin
    Cruz, Breno Dantas
    Tilevich, Eli
    COMPUTER COMMUNICATIONS, 2023, 206 : 133 - 151