Hasan, Maaz (2019) A Hybrid Real-Time Intrusion Detection System for an Internet of Things Environment with Signature and Anomaly Based Intrusion detection. Masters thesis, Dublin, National College of Ireland.
|
PDF (Master of Science)
Download (679kB) | Preview |
|
|
PDF (Configuration manual)
Download (410kB) | Preview |
Abstract
Intrusion detection systems play important role in real world applications. Every organization or government that uses any sort of networking and information systems need protection from various kinds of intrusions. Many existing intrusion detection systems provide very highly verbose output and it is not easier for administrators to identify the issues immediately. With the Artificial Intelligence (AI) techniques with underlying Machine Learning (ML) algorithms, there is scope of developing IDS based on AI. In this project, a hybrid IDS is developed using machine learning approaches. It combines Random Forest classification and K-Means clustering. This will use both misuse detection and anomaly detection for improving performance of the IDS. These algorithms are evaluated for the four categories of attacks based on precision, recall, F1-score, false-alarm-rate, and detection-rate. The proposed IDS is evaluated with NSL-KDD dataset which is highly optimized for intrusion detection research. The results of experiments showed that the hybrid IDS perform well in terms of detection rate and other metrics.
Actions (login required)
![]() |
View Item |