Scholarlink

This study proposes a privacy-preserving digital economy model optimization framework based on federated learning, addressing the challenge of balancing data privacy and economic efficiency in distributed systems. The model integrates differential privacy and secure multi-party computation to protect client data, while optimizing global loss and utility through a client-server architecture. Experimental results demonstrate that increasing the privacy budget from 0.5 to 10 enhances model accuracy from 0.68 to 0.90 and reduces global loss from 0.42 to 0.20, highlighting the trade-off between privacy constraints and performance. The framework effectively supports applications in e-commerce and financial networks by leveraging distributed training and noise-perturbed gradient updates, providing a robust solution for secure data collaboration in the digital economy.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Optimization of Privacy-Preserving Digital Economic Models Under the Federated Learning Framework

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Jieqiong Gong,
  • Xiaoqing Liang

摘要

This study proposes a privacy-preserving digital economy model optimization framework based on federated learning, addressing the challenge of balancing data privacy and economic efficiency in distributed systems. The model integrates differential privacy and secure multi-party computation to protect client data, while optimizing global loss and utility through a client-server architecture. Experimental results demonstrate that increasing the privacy budget from 0.5 to 10 enhances model accuracy from 0.68 to 0.90 and reduces global loss from 0.42 to 0.20, highlighting the trade-off between privacy constraints and performance. The framework effectively supports applications in e-commerce and financial networks by leveraging distributed training and noise-perturbed gradient updates, providing a robust solution for secure data collaboration in the digital economy.