Intrusion Detection Using the WEKA Machine Learning Tool
Date
2021-12-20
Authors
Sadiq, Ali
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The internet has become essential for data sharing, so it must be secure. Data encryption
and authentication are not sufficient for internet security, and firewalls are unable to
detect fragmented malicious packets. Further, attackers are constantly modifying their
techniques and tools to produce devastating consequences such as productivity loss, financial loss and brand damage. Thus, it has become crucial to implement an efficient
intrusion detection system, which is a very difficult challenge. The CSE-CIC-IDS2018
dataset consists of 14 network attacks and benign traffic with 80 features. In this report,
data for seven network attacks and benign data are used. Principal Components Analysis
(PCA) is employed to extract the ten most relevant features. The Machine Learning (ML)
algorithms studied are Multi-Layer Perceptron (MLP), Support Vector Machine (SVM),
K-Nearest Neighbor (KNN), Random Forest (RF) and Bayesian Network (BayesNet).
The experiments are performed using the Waikato Environment for Knowledge Analysis
(WEKA) tool with five-fold and ten-fold cross-validation. The performance measures
employed are accuracy, precision, recall, F-measure, and execution time. The results obtained show that RF outperforms the other algorithms in accuracy, precision, recall, and
F-measure.