Network analytics techniques: Next-generation abnormal behavior detection system based on big data analytics and deep learning techniques
Wang, Huiqiang
Wang, Huiqiang
Type
Supervisor
Subject
Date
2026-01-01
Collections
Research Projects
Organizational Units
Journal Issue
Abstract
Abnormal behavior detection, in the new era of big data, presents a profound challenge in the field of cybersecurity. Traditionally, cybersecurity relied on a paradigm of prevention and ad hoc detection of intrusions resulting from attacks on networks. However, the landscape has evolved significantly with the advent of large-scale data and networks, introducing new complexities and vulnerabilities. This chapter is devoted to addressing the evolving problem of intrusion detection in the wake of unprecedented data volumes and expansive networks. It underscores the intrinsic connections and significance of big data analytics, deep learning, and cybersecurity principles in this context. Our primary focus lies in the practical application of advanced autoencoder variants, including Autoencoders, Sparse Autoencoders, Contractive Autoencoders, and Convolutional Autoencoders (CAEs) within the distributed Spark framework. The main aim of this chapter is to empower readers with a comprehensive understanding of mathematical modeling and algorithmic approaches in the domain of big data analytics. With a particular emphasis on large-scale abnormal behavior detection, we provide practical insights, real-world illustrations, and comprehensive perspective. By exploring these advanced techniques and their integration with distributed computing platforms, we seek to contribute to the ongoing evolution of abnormal detection strategies, making them better suited to the challenges of the new age of big data and large-scale networks. Furthermore, we outline methodological practical steps for implementing this innovative solution, emphasizing the synergy between mathematical modeling and real-world application. By highlighting the novel contributions and practical steps for implementation, this chapter offers a valuable resource for researchers and practitioners seeking to navigate the intersection of big data, deep learning, and cybersecurity effectively.
Department
Publisher
Sponsor
na
Copyright
Book title
Mathematical Modeling for Big Data Analytics