A Survey On Feature Selection And Classifiers In Machine Learning For The Intrusion Detection Systems

Abstract

The security of a computer system and the network is getting into a serious issue because of smart and intelligent hackers. An increase in more usage of the internet leads to the Detection of threats and attacks to be considered crucial to protect computer systems from many attacks though firewalls are installed. This made researchers insights into various classification and machine learning algorithms to build an efficient intrusion detection system. Intrusion Detection Systems combined with other technologies improves detection and prediction rates.  Machine Learning algorithms automatically discover useful information from massive datasets. Though various techniques are available, every day some new born attacks are identified which highlights the need for robust techniques to identify attacks. Different researchers proposed different algorithms in different categories for IDS. NSL KDDCup99 is the dataset used by most of the researchers to work in the field of the intrusion detection system.  This paper presents a comprehensive review of Intrusion Detection Systems, DataSets and various researches related to Machine Learning based IDS by researchers in the last decade.