A Study on Artificial Intelligence and Machine Learning Based Intrusion Detection Systems for Detecting Cyber Attacks

Abstract

Intrusion detection Systems (IDS) is software and hardware that examines the network traffic and tries to find possible attacks and intrusions. The usage of technologies and services has increased tremendously on the internet and at the same time the numbers of cyber attacks in various forms and means have also increased.  Many techniques and methods are available to detect malicious attacks in the networked environment.  Each of those methods has some shortfalls and failures to identify complex and dynamic attacks. Artificial Intelligence (AI) and Machine Learning (ML) techniques were used to overcome the issues with the existing intrusion detection systems. This work focuses on some of the AI and ML based for IDS and those methods were evaluated using KDD99 dataset. It is found that the decision tree algorithm shows efficient results in terms of precision, F-Score, False alarm rate and accuracy .