Comparative Analysis of Different Machine Learning Algorithms on Intrusion Detection System

Abstract

Intrusion detection is a very important issue in the 21st century where people are using a greater number of computers which are connected over different networks. Even after previously proposed various techniques like packet filtering, application gateway, circuit-level gateway, proxy server, encryption, to stop intruders to get into the system through a network, they have somehow managed to find the way to get to the system and steal data. This issue has been certainly come down after machine learning came into the picture. This paper reviews different machine learning algorithms on intrusion detection system and also provides directions for further research. We have considered NSL-KDD, KDD-CUP99 dataset for the evaluation, which is the best dataset considered by the researchers for the intrusion detection system.