Investigating Network Intrusion Detection Datasets Using Machine Learning

Abstract

There's been a series of datasets with regards to network intrusion detection in recent years, and a significant number of studies has also been carried out using these datasets. In this paper we aim to explore these datasets while also showing the capability of the proposed model to accurately detect and classify network intrusions. This paper presents a deep learning based model for network intrusion as well as a comparative analysis of the performance of three major network intrusion datasets using the proposed model. Results showed the model to perform best for NSL-KDD, followed by UNSW-NB15 and CSE-CIC-IDS2018 respectively. Model accuracy achieved for these datasets were NSL-KDD (97.89%), UNSW-NB15 (89.99%), and CSE-CIC-IDS2018 (76.47%) was achieved.