The Analysis of KDD-Parameters to Develop an Intrusion Detection System Based on Neural Network
Ilhame El Farissi, Sara Chadli, Mohamed Emharraf, Mohammed Farooq Saber · Lecture notes in electrical engineering · 2016
The intrusion Detection System (IDS) is designed to protect a computer or a network by detecting malicious attempts to storm the system. Therefore, it is important to develop a flexible IDS which is able to detect attacks with best performance. In recent researches, most of IDS are based on neural network and alimented by KDD data. Which means that the neural networks inputs correspond to the KDD-parameters. However, some of KDD-parameters are meaningless and can increase the detection rate. In order to optimize the IDS performance, it is primordial to exploit uniquely the most important and crucial parameters. In fact, there are three categories of KDD attributes; the basic attributes, the parameters relating to content and the time-based ones. The study carried out in this work consists on selecting the most efficient parameters in intrusion detection by demonstrating their utility and performance and neglecting the meaningless attributes. It has to be emphasized that MATLAB tool has been used to develop and put into practice the IDS in question.