Application of Personal Information Privacy Protection Based on Machine Learning Algorithm

Area of Science:

• Computer Science
• Data Science
• Cybersecurity

Background:

• Personal information privacy is a growing concern in the digital age.
• Joint machine learning platforms offer collaborative model training but face privacy challenges.
• Existing privacy protection methods may be insufficient against sophisticated attacks.

Purpose of the Study:

• To develop and evaluate a machine learning algorithm for robust personal information privacy protection.
• To design a privacy-preserving joint machine learning architecture.
• To introduce a data homomorphic verification mechanism for enhanced security.

Main Methods:

• Definition and implementation of a novel machine learning algorithm.
• Design of a joint machine learning platform architecture with privacy considerations.
• Development of a data homomorphic verification mechanism to prevent parameter tampering.
• Comparative experimental analysis of model performance.

Main Results:

• The proposed algorithm exhibits fast convergence.
• Average error rate reduced by 4.17% compared to traditional algorithms.
• Achieved an accuracy of 95.37%, an 8.76% improvement over previous methods.
• Demonstrated effectiveness in preventing global parameter tampering.

Conclusions:

• The developed machine learning algorithm provides a secure and reliable environment for personal information privacy.
• The approach offers a significant advancement in protecting data owners' privacy.
• This work provides valuable insights for future research in privacy protection technologies.