Work

An Improved Explainable Artificial Intelligence for Intrusion Detection System

Year: 2024
Type: article
Abstract: Cybersecurity professionals rely heavily on Intrusion Detection Systems (IDS) to identify as well as stop potential dangers. Networks may be better protected with the use of IDS. A variety of Machine Learning (ML) approaches are aimed at the development of successful IDSs. Ensemble methods in ML have a history of successful learning. This research proposes a cutting-edge IDS using ensemble methods of ML. This research used preprocessing data after the CICIDS-2017 dataset to enhance classification accuracy and suppress false positives. Using ML methods including Logistic Regression, XGBoost (XGB) ID classifiers, along with Light Gradient Boosting Machine (LGBM) classifiers, proposes an IDS. An ensemble technique classifier was applied after these models were trained, and accuracy was obtained. The suggested model also includes the Explainable Artificial Intelligence (XAI) algorithm Local interpretable model-agnostic explanation (LIME), which makes the for reliable ID easier to understand and explain. The XAI LIME is faster, more responsive, and easier to explain.
Institution: Jazan University
Language: English
Open Access status: diamond
OpenAlex ID: W7217565237
FWCI: 0.749
Cites: 33
Cited by: 7
Related to: 10