Machine Learning for Intrusion Detection: A Reproducible Baseline is All You Need

Main Article Content

Salma A. Walli
Karam Sallam

Abstract

Ensuring Responsible AI practices is paramount in the advancement of systems founded upon machine learning (ML) principles, particularly in sensitive domains like intrusion detection within cybersecurity. A fundamental aspect of Responsible AI is reproducibility, which guarantees the reliability and transparency of research outcomes. In this paper, we address the critical challenge of establishing reproducible for intrusion detection utilizing ML techniques. Leveraging the NSL-KDD dataset and the Edge-IIoTset, we carry out extensive experiments to evaluate the efficacy of our approach. Our study prioritizes meticulous experiment design and careful implementation setups, aligning with the principles of Responsible AI. Through rigorous experimentation and insightful discussions, we underscore the importance of reproducibility as a cornerstone in ensuring the resilience and reliability of intrusion detection systems. Our findings offer valuable insights for researchers and practitioners striving to develop Responsible AI solutions in cybersecurity and beyond. The source code is publicly accessible at https://github.com/Salma-00/Machine-Learning-for-Intrusion-Detection

Downloads

Download data is not yet available.

Article Details

How to Cite
A. Walli, S. and Sallam, K. (2024) “Machine Learning for Intrusion Detection: A Reproducible Baseline is All You Need”, Sustainable Machine Intelligence Journal, 7, pp. (3):1–29. doi:10.61356/SMIJ.2024.77103.
Section
Original Article

How to Cite

A. Walli, S. and Sallam, K. (2024) “Machine Learning for Intrusion Detection: A Reproducible Baseline is All You Need”, Sustainable Machine Intelligence Journal, 7, pp. (3):1–29. doi:10.61356/SMIJ.2024.77103.