Network Intrusion Detection using Decision Tree and Random Forest

With the growth in technology and increase in the usage of the same, we have seen that how Network Intrusion Detection has emerged and have also became a most important aspect in the area of research. The Intrusion Detection System broadly classifies the attack as normal or hostile attack based on the activities of the users. An Intrusion Detection System deals with Traffic data of networks. It is non-linear and can deal with complicated problems. In the past researches, many Intrusion Detection system was proposed with different accuracy levels. Even though there is no model that can precisely detect or predict an attack. Therefore, it is important to develop a robust and effective Intrusion Detection Model. In this paper a Network Intrusion Detection System is developed using Decision Tree and Random Forest classifier. These techniques give us a better accuracy and performs pretty well when compared with some other traditional classifier for classifications of attack effectively. The NSL-KDD dataset have been used to conduct our experiments and evaluate the performance of our model. The final results shows that our proposed method is efficient enough to give a low false alarm rate and high detection rate.