A Machine Learning Approach for Network Intrusion Detection System

With the increase in usage of networking technology and to the advancement of Internet over the last decade, network security has turned out to be one of the important areas of research. Therefore, Intrusion Detection (ID) becomes important and challenging security problem and is widely studied to achieve overall network security success. Several techniques came into existence to detect intrusions based on Machine Learning. Aim of Intrusion Detection System (IDS) is to put checks on attacks and provide desirable support for defense management along with information about the intrusion. Several ID approaches are proposed so far to predict malicious traffic from the network. In this paper, a hybrid approach for the network traffic classification is proposed where Recursive Feature Elimination will be used for selecting feature and Decision Tree will classify the data based on extracted features. The accuracy observed on test data of the classifier is 99%.