Network Intrusion Detection System Using Decision Tree and KNN Algorithm

The complexity of attacks is increasing, making it increasingly difficult to effectively discover breaches. A network intrusion detection system is needed to handle the aforementioned problem. An automated software program called a network intrusion detection system alerts administrators when someone attempts to compromise the system by participating in hazardous behaviors. A system or network is protected from harmful invasions by hardware and software firewalls. They function like filters, removing any data that could put the system or network at risk. We aimed to design the best intrusion detection system possible so that it could distinguish between "normal" and "attacked" categories of network data with high accuracy. The best accuracy can be obtained by using soft computing techniques like Decision Trees and KNN. We tested a variety of techniques, such as preprocessing the data, feature selection, principal component analysis (PCA) reduction, standardization, and normalization, in order to increase our model's accuracy scores. We also assess the two approaches' results according to their accuracy in order to decide which way is more effective.