An intelligent intrusion detection system

With the introduction of emerging technologies cybersecurity has become an inherited and amplified problem. New technologies bring significant developments but also come with new challenges in the cybersecurity area. The fight against malicious attacks is an everyday battle for every company. Challenges brought by security breaches can be devastating for a company and sometimes bring un-survivable circumstances. In this paper, we propose a novel two-stage intelligent intrusion detection system (IDS) to detect and protect from such malicious attacks. Intrusion Detection Systems are feasible solutions for cybersecurity problems, but they come with implementation challenges. Anomaly based IDS usually have a high rate of false positives (FP) and they require considerable computational requirements. The approach proposed in this paper consists of a two-stage architecture based on machine learning algorithms. In the first stage, the IDS uses K-Means to detect attacks and the second stage uses supervised learning to classify such attacks and eliminate the number of false positives. The implementation of this approach results in a computationally efficient IDS able to detect and classify attacks at a 99.97% accuracy while lowering the number of false positives to 0. The paper also evaluates the performance results and compares them with other relevant research papers. The performance of this proposed IDS is superior to the current state of the art.