Privacy-Preserving Neural Networks for Collaborative Cybersecurity

EasyChair Preprint no. 14014

Abstract

In the era of increasing cyber threats and attacks, collaborative cybersecurity has emerged as a powerful approach to enhance the collective defense against malicious activities. However, sharing sensitive data between different organizations raises concerns about privacy and data protection. This paper introduces privacy-preserving neural networks as a solution to address these concerns and enable secure collaboration in cybersecurity.

The proposed approach leverages advanced cryptographic techniques and secure multi-party computation to train neural networks on distributed datasets without compromising the privacy of individual organizations. By encrypting the data and performing computations on encrypted data, the privacy-preserving neural networks ensure that sensitive information remains protected throughout the collaboration process.

Furthermore, this paper presents a detailed analysis of the performance and effectiveness of privacy-preserving neural networks in the context of collaborative cybersecurity. Experimental results demonstrate that the proposed approach achieves comparable accuracy to traditional neural networks while preserving the privacy of the participating organizations.

Keyphrases: Malicious, networks, neural

@Booklet{EasyChair:14014,
  author = {Kaledio Potter and Dylan Stilinki and Ralph Shad},
  title = {Privacy-Preserving Neural Networks for Collaborative Cybersecurity},
  howpublished = {EasyChair Preprint no. 14014},

  year = {EasyChair, 2024}}