Netwrok-Intrusion-Detection-system

The IDS we will be developing is a network-based IDS that is programmed to detect any misuse of the network resources (misuse detection) i.e., it detects malicious packets flowing in a network. We will be building a classifier using Decision Trees. They help in increasing the accuracy at which intrusions are detected. Before building the classifier, we will be required to select the most optimal features using Feature selection. We will be employing the concept of Recursive Feature Elimination (RFE). Because of RFE the attacks are detected more efficiently even in highly congested networks. It also leads to a lesser number of false positives and therefore a lower rate of false alarms. Also, The time taken to detect attacks can be cut down significantly by using databases for storage and the concept of Dynamic multi-boosting. The proposed network intrusion detection system can classify packets in real-time based on the packets collected from the network flow.