SIMULATING AND ANALYZING SIGNATURE-BASED NETWORK INTRUSION DETECTION SYSTEMS USING MACHINE LEARNING TECHNIQUES

Fuente: Zenodo
Saved in:
Bibliographic Details
Main Authors: Issah, Peter Paul, Ganyo, Ransford, Salam, Abdul Gaddafi
Format: Recurso digital
Language:English
Published: Zenodo 2022
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866902291286589440
author Issah, Peter Paul
Ganyo, Ransford
Salam, Abdul Gaddafi
author_facet Issah, Peter Paul
Ganyo, Ransford
Salam, Abdul Gaddafi
contents <ul> <li><strong>Objective:</strong> The main objective of this study was to develop and evaluate intrusion detection models using machine learning algorithms to enhance security in data sharing over the internet.</li> <li><strong>Methods:</strong> The research utilized Weka Data Mining Software to create intrusion detection models based on the UNSW-NB15 dataset, which includes 44 attributes and 9 different types of attacks. Principal Components Analysis (PCA) was applied to reduce the number of features and improve detection speed. The machine learning methods evaluated were K-Nearest Neighbor (KNN), Random Forest (RF), Bayesian Network (BayesNet), and Decision Tree (J48).</li> <li><strong>Results:</strong> Experiments were conducted using the WEKA tool with ten-fold cross-validation, and performance was assessed based on execution time, F-measure, recall, accuracy, and precision. The Decision Tree (J48) algorithm outperformed other algorithms in terms of accuracy, precision, recall, and F-measure, while the Bayesian Network (BayesNet) algorithm produced the least accurate results.</li> <li><strong>Significance:</strong> This research provides a robust approach to enhancing intrusion detection systems, which is crucial for maintaining security and mitigating risks associated with internet data sharing. The findings indicate that the J48 algorithm is particularly effective for this purpose, offering a reliable tool for detecting intrusions and ensuring data integrity.</li> <li><strong>Keywords:</strong> Intrusion Detection, Machine Learning, Principal Components Analysis, UNSW-NB15 Dataset, WEKA, Security, Data Mining</li> </ul>
format Recurso digital
id zenodo_https___doi_org_10_5281_zenodo_12695351
institution Zenodo
language eng
publishDate 2022
publisher Zenodo
record_format zenodo
spellingShingle SIMULATING AND ANALYZING SIGNATURE-BASED NETWORK INTRUSION DETECTION SYSTEMS USING MACHINE LEARNING TECHNIQUES
Issah, Peter Paul
Ganyo, Ransford
Salam, Abdul Gaddafi
Intrusion Detection, Machine Learning, Principal Components Analysis, UNSW-NB15 Dataset, WEKA, Security, Data Mining
<ul> <li><strong>Objective:</strong> The main objective of this study was to develop and evaluate intrusion detection models using machine learning algorithms to enhance security in data sharing over the internet.</li> <li><strong>Methods:</strong> The research utilized Weka Data Mining Software to create intrusion detection models based on the UNSW-NB15 dataset, which includes 44 attributes and 9 different types of attacks. Principal Components Analysis (PCA) was applied to reduce the number of features and improve detection speed. The machine learning methods evaluated were K-Nearest Neighbor (KNN), Random Forest (RF), Bayesian Network (BayesNet), and Decision Tree (J48).</li> <li><strong>Results:</strong> Experiments were conducted using the WEKA tool with ten-fold cross-validation, and performance was assessed based on execution time, F-measure, recall, accuracy, and precision. The Decision Tree (J48) algorithm outperformed other algorithms in terms of accuracy, precision, recall, and F-measure, while the Bayesian Network (BayesNet) algorithm produced the least accurate results.</li> <li><strong>Significance:</strong> This research provides a robust approach to enhancing intrusion detection systems, which is crucial for maintaining security and mitigating risks associated with internet data sharing. The findings indicate that the J48 algorithm is particularly effective for this purpose, offering a reliable tool for detecting intrusions and ensuring data integrity.</li> <li><strong>Keywords:</strong> Intrusion Detection, Machine Learning, Principal Components Analysis, UNSW-NB15 Dataset, WEKA, Security, Data Mining</li> </ul>
title SIMULATING AND ANALYZING SIGNATURE-BASED NETWORK INTRUSION DETECTION SYSTEMS USING MACHINE LEARNING TECHNIQUES
topic Intrusion Detection, Machine Learning, Principal Components Analysis, UNSW-NB15 Dataset, WEKA, Security, Data Mining
url https://doi.org/10.5281/zenodo.12695351