Abstract

The escalating significance of cybersecurity, due to IoT’s growth, demands robust security. As cyberattacks increase, machine learning-based network intrusion detection systems (NIDS) provide an effective countermeasure. This paper conducts experiments to optimize an NIDS pipeline using three artificial neural network (ANN) paradigms, demonstrating the importance of optimization and addressing computational time misconceptions. It assesses realistic datasets and compares performance metrics and execution times. Our main contribution is evaluat- ing data processing pipelines for ANN application in NIDS, and benchmarking processing ap- proaches’ influence on advanced neural-network methods.

Recommended Citation

Pawlicki, M. (2023). Strengths And Weaknesses of Deep, Convolutional and Recurrent Neural Networks in Network Intrusion Detection Deployments. In A. R. da Silva, M. M. da Silva, J. Estima, C. Barry, M. Lang, H. Linger, & C. Schneider (Eds.), Information Systems Development, Organizational Aspects and Societal Trends (ISD2023 Proceedings). Lisbon, Portugal: Instituto Superior Técnico. ISBN: 978-989-33-5509-1. https://doi.org/10.62036/ISD.2023.54

Paper Type

Short Paper

DOI

10.62036/ISD.2023.54

Share

COinS
 

Strengths And Weaknesses of Deep, Convolutional and Recurrent Neural Networks in Network Intrusion Detection Deployments

The escalating significance of cybersecurity, due to IoT’s growth, demands robust security. As cyberattacks increase, machine learning-based network intrusion detection systems (NIDS) provide an effective countermeasure. This paper conducts experiments to optimize an NIDS pipeline using three artificial neural network (ANN) paradigms, demonstrating the importance of optimization and addressing computational time misconceptions. It assesses realistic datasets and compares performance metrics and execution times. Our main contribution is evaluat- ing data processing pipelines for ANN application in NIDS, and benchmarking processing ap- proaches’ influence on advanced neural-network methods.