False alarm reduction for Network Intrusion Detection System by using Decision Tree classifier


Nowadays, Network security is one of the challenging issues with the rapid growth in information technology, this subject leading people to become increasingly aware of the threats to personal privacy through computer crime. Therefore, there is important to create intrusion detection system to detect malicious activities and various attacks on the internet with elevated detection rate and minimal false positive alarm. This paper proposed Network Intrusion Detection system using Decision Tree algorithm. To detect and classify attacks into four categories (DOS, Probe, R2L, U2R). The KDDcup99 dataset has been used to evaluate the activity of proposition system. The experimental results showed that the proposed system provides better results with high detection rate in experiment 1 (99.95%), experiment 2 (97.8%) and low false alarm rate in experiment 1 (0.05%), experiment 2 (2.2%).