Improved Intrusion Detection System (IDS) performance using Machine Learning: A Comparative Study of Single Classifier and Ensemble Learning

Unlike conventional firewalls, intrusion detection systems (IDS) are acknowledged as a very significant and leading strategy for detecting hazardous behaviors on computer networks, and they are rapidly replacing them. Unlike traditional firewalls, IDS employs analytical methodology such as data mining and machine learning techniques to intelligently identify threats, whereas traditional firewalls do not. Ensemble learning has spurred progress in machine learning and pattern classification research in recent decades, with results that surpass single classifiers. Research was carried as part of this study to improve the accuracy rate of an intrusion detection system. To evaluate which technique worked best, a singular classifier was used to identify given data in order to obtain the highly accurate results, which were then evaluated to the findings of ensemble learning and feature selection. Using ensemble learning, the goal is to obtain the most accuracy out of a single classifier. The results are produced from the confusion matrix value, and they will be evaluated through making a comparison acquired using the two different methods stated above. The researchers were able to get an accuracy score of 77.4% for a single classifier and 96.8% for ensemble learning.