Proceedings of the 2nd International Conference on Financial Innovation, FinTech and Information Technology, FFIT 2023, July 7–9, 2023, Chongqing, China

Research Article

Research on the Application of Deep Learning-based Differential Privacy Protection Models in Financial Big Data

Download147 downloads
  • @INPROCEEDINGS{10.4108/eai.7-7-2023.2338057,
        author={Yanjun  Ma and Rui  Lu and Yu  Zhang},
        title={Research on the Application of Deep Learning-based Differential Privacy Protection Models in Financial Big Data},
        proceedings={Proceedings of the 2nd International Conference on Financial Innovation, FinTech and Information Technology, FFIT 2023, July 7--9, 2023, Chongqing, China},
        publisher={EAI},
        proceedings_a={FFIT},
        year={2023},
        month={10},
        keywords={deep learning differential privacy privacy protection financial big data},
        doi={10.4108/eai.7-7-2023.2338057}
    }
    
  • Yanjun Ma
    Rui Lu
    Yu Zhang
    Year: 2023
    Research on the Application of Deep Learning-based Differential Privacy Protection Models in Financial Big Data
    FFIT
    EAI
    DOI: 10.4108/eai.7-7-2023.2338057
Yanjun Ma1, Rui Lu1, Yu Zhang2,*
  • 1: Liaoning Police College
  • 2: Dalian Medical University
*Contact email: zhangyu02xinben@163.com

Abstract

With the digital transformation of the financial industry and the widespread application of big data, privacy protection has become a critical challenge. The differential privacy protection model based on deep learning preserves the privacy of financial big data while maintaining data availability and accuracy. This paper investigates and analyses the application of differential privacy protection models based on deep learning in the field of financial big data. Firstly, this paper introduces the basic concepts and principles of differential privacy, as well as the significance of deep learning in the financial domain. Next, the working principles and main methods of the differential privacy protection model based on deep learning are elaborated, including privacy budget allocation, noise injection, and model optimization techniques. Finally, this paper explores the application of the differential privacy protection model based on deep learning in financial big data analysis. It is noted that these models can achieve efficient and accurate data analysis and prediction while protecting user privacy. Specific applications include individual credit assessment, fraud detection, anti-fraud measures, among others. The paper also discusses the advantages and potential challenges of the models and proposes future research directions.