Development of Cyber Threat Early Detection System Using Distributed Machine Learning Algorithms

Abstract

This research aims to develop a distributed machine learning-based system for the early detection of cyber threats. With the rise of cyberattacks targeting critical sectors such as education, government, and healthcare, traditional intrusion detection systems have become less effective at identifying novel threats. To address this, the study introduces a federated learning approach, which allows machine learning models to be trained across distributed nodes while maintaining data privacy. The system architecture integrates various nodes in a collaborative manner, enabling real-time detection of cyber threats with improved efficiency and data security. The study evaluates the system's performance using real-world datasets and compares the federated approach with centralized models, achieving competitive accuracy and privacy benefits. The findings highlight the importance of system speed, education level, and the role of federated learning in improving cybersecurity. This research contributes to the development of more adaptive, scalable, and privacy-preserving security systems in the context of modern network infrastructures