Multi-Attacks Detection in Distributed System using Machine Learning

Abstract

Intrusion compromises a computer by breaking its security and thereby the computer enters into an insecure state. If such an event takes place, the computer becomes vulnerable to several attacks. These attacks aim to obtain information about the target computer and the information so obtained can be used to conduct fraudulent activities. It is difficult to prevent an intrusion into the system. However, if these computer intrusions are detected in time, the administrator can be informed and necessary actions can be taken at early stages. Previous Intrusion detection system (IDS) utilized several features to detect various malicious activities. However, these IDS methods only detect specific attack. They fail when the attacks are combined. For this purpose, we propose an Intrusion Detection System in distributed environment to mitigate the individual and combination routing attacks. This paper explains the method we used to generate such a system. Our proposed system of Intrusion Detection uses feature selection techniques to determine significant features, along with the best classification method will distinguish between an attack and non-attack. We aim to increase detection accuracy and reduce false alarm rate. NSLKDD dataset has been used to train our model. The paper also explains related work done in this field and briefly explains the network attacks and the dataset.