#Intelligent Secure Systems ##Using a Joone Neural Network for Intrusion Detection Intrusion Detection Systems (IDS) are designed to classify activity in order to differentiate between legitimate and illegitimate use of a computer or network. An IDS may monitor any type of activity, but a common type of activity to monitor is network traffic. An intrusion detection system which monitors network activity is called a Network-based Intrusion Detection System, or NIDS. Network traffic is relatively easy to capture and any intrusion from a remote machine should be present in a comprehensive log of the network activity. The problem that an NIDS attempts to solve is finding the traffic which should be classified as illegitimate activity and reporting it appropriately.
Finding illegitimate activity is a very difficult problem in most cases. Since a network is used for many diverse tasks, many times illegitimate traffic very closely resembles legitimate traffic. In this project we examine the application of neural networks to the problem of classification of traffic.
Final Paper: PDF