Authors:
P. Dhinakaran, M.A. Thinesh, Mykhailo Paslavskyi

Addresses:
1,2Department of Computer Science and Engineering, SRM Institute of Science and Technology, Ramapuram, Chennai, Tamil Nadu, India. 3Department of Computer Science, Ukrainian National Forestry University, Lviv City, Ukraine.  pp4417@srmist.edu.in1, tm9045@srmist.edu.in2, mykhailo.paslavskyi@nltu.edu.ua3

Abstract:

Intrusions can interrupt network operations, steal critical data, and gain unauthorised access to network resources. Detecting and avoiding network breaches is critical to maintaining network security. Intrusion Detection Systems (IDS) and Intrusion Prevention Systems (IPS) monitor network traffic for unusual behavior and respond to threats. Ensemble learning techniques are used to identify cyber intrusions to improve detection accuracy and resilience. The KDD CUP 99 dataset contains a variety of variables collected from cyber network traffic, including duration, byte transfer rates, protocol kinds, and more. Pre-processing stages include categorical variable encoding and feature selection to prepare the dataset for modeling. As a result, Random Forest classifiers serve as foundation learners for bagging and stacking ensemble techniques. The performance of these ensemble models is evaluated using a variety of measures, including accuracy, precision, recall, and F1 score. Furthermore, visualization approaches such as confusion matrices help to analyze categorization performance across different cyber incursion types. This paper uses empirical assessment to demonstrate the usefulness of ensemble learning paradigms in cyber intrusion detection, highlighting their value in improving network security against various cyber threats.

Keywords: Intrusion Detection Systems; Intrusion Prevention Systems; Ensemble Learning Techniques; Random Forest Classifiers; Categorical Variable Encoding; Feature Selection; Performance Evaluation; Cyber Incursion Types.

Received on: 02/04/2023, Revised on: 16/07/2023, Accepted on: 05/10/2023, Published on: 22/12/2023

FMDB Transactions on Sustainable Computer Letters, 2023 Vol. 1 No. 4, Pages: 210-227

• Views : 219
• Downloads : 27

Download PDF