A Comparative Study of Machine Learning Algorithms Used for Network Intrusion Detection

• Author(s): Ijegwa David Acheme ; Adebanjo Adeshina Wasiu
• Paper ID: 1706057
• Page: 494-500
• Published Date: 30-07-2024
• Published In: Iconic Research And Engineering Journals
• Publisher: IRE Journals
• e-ISSN: 2456-8880
• Volume/Issue: Volume 8 Issue 1 July-2024

Abstract

Across different sectors of human endeavors such as health, aviation, agriculture, education, and finance, institutions and organizations are increasingly embracing ICT infrastructures, relying more on computers and cyber resources for their daily operations. However, this growing dependence on cyber systems has led to a corresponding rise in cyber-attacks. Therefore, there's a pressing need to develop robust countermeasures to safeguard confidential information and ensure its availability. Among the many techniques employed by attackers to breach computer network security, intrusion stands out as one of the most common significant attack type. Numerous research endeavors have been dedicated to developing intrusion detection systems (IDS) to address this challenge. The focus of this study is the exploration of selected machine learning techniques that have been reported for IDS development. To accomplish this, the research builds a predictive machine learning model using four popular algorithms (Logistic Regression, Random Forest, Decision trees, and Naïve Bayes) for the detection and prediction of suspicious connections. This was achieved through the analysis of the KDD Cup 1999 dataset, wherein machine learning algorithms are employed to identify patterns and anomalies, which can enable business owners to deploy preemptive measures against potential security breaches. Subsequently, the performances of these algorithms are evaluated and ranked based on their prediction accuracy and other established performance metrics. The results in terms of prediction accuracy show that the Random Forest algorithm performed best, followed by the decision tree, then Logistic regression and finally, naïve bayes with the least accuracy.

Keywords

Classification Algorithms, Cybersecurity, Intrusion Detection Systems

Citations

IRE Journals:
Ijegwa David Acheme , Adebanjo Adeshina Wasiu "A Comparative Study of Machine Learning Algorithms Used for Network Intrusion Detection" Iconic Research And Engineering Journals Volume 8 Issue 1 2024 Page 494-500

IEEE:
Ijegwa David Acheme , Adebanjo Adeshina Wasiu "A Comparative Study of Machine Learning Algorithms Used for Network Intrusion Detection" Iconic Research And Engineering Journals, 8(1)