Intelligent Intrusion Detection System Using a Committee of Experts

Intrusion detection plays an important role in today's computer and communication technology. It is of a paramount importance to design a time efficient and accurate Intrusion Detection System (IDS). In this paper, an Intelligent Intrusion Detection System (IIDS) is developed using the combination of several neural experts forming a Committee of Experts, used to detect and prevent various anomalies. The input to the system is given by the feature extraction of various historic data present in the central database and making a suitable training data classifier. Furthermore, each expert within the committee is assigned a different granularity of data, because the training set is based on its fundamental properties. The experts are designed in such a way that they make accurate predictions using the data. The system takes the necessary steps in identification of the anomaly and was found to have a high attack success detection rate.