Intrusion Detection System on Nowaday's Attack using Ensemble Learning

Fajar Henri Erasmus Ndolu (Universitas Indonesia)
Ruki Harwahyu (Universitas Indonesia)

International Journal of New Media Technology
Vol 10 No 1 (2023): IJNMT

Attacks on computer networks are becoming more and more widespread nowadays, making this an important issue that must be considered . These attacks can be detected with the Intrusion Detection System (IDS). However, at this time there are new attacks that have not been detected by IDS. Therefore, ensemble learning is used. This research we used Random Forest algorithm for attack detection as an increase in the ability of IDS to detect cyber attacks. The use of the CSE-CIC-IDS2018 dataset is used in this research as a current representative dataset for cyber attack detection. The results of this study we get a binary classification accuracy of 99.6856% and an f1-score of 99.5803% and a multiclass classification accuracy of 99.6944 and an f1-score of 97.8032% with a data ratio ratio dataset of 3:1 normal class to attack class.