Intrusion detection using the adaptive clustering and optimization based classification approach

Abstract

A cyber-physical system is a mechanism that is controlled by computer-based algorithms, tightly integrated with the Internet and its users. Cyber security is the protection of internet-connected systems, including hardware, software and data, from cyber attacks. Further, Intrusion detection system has emerged as one of the important security process in the recent years, as it helps intrusion detection and prevention systems are primarily focused on identifying possible incidents, logging information about them, reporting attempts and malicious attacks. Here, the security is done by developing intrusion detection. The proposed schemes are named as I-AHSDT: Intrusion Detection using Adaptive Dynamic Directive Operative Fractional Lion clustering and Hyperbolic Secantbased Decision Tree Classifier and Crow-AFL: Crow based adaptive fractional lion optimization approach for the intrusion detection algorithms. As the primary contribution, a Crow based Adaptive Fractional Lion optimization approach the proposed IDS clusters the database into several groups with the Crow-AFL and detects the presence of intrusion in the clusters with the use of the HSDT classifier. Final contribution, I-AHSDT, which is the combination of the Adaptive Dynamic Directive Operative Fractional Lion clustering (ADDOFL) and Hyperbolic Secant-based Decision Tree classifier (HSDT). The proposed method inherits the adaptive and the global optimal nature of the Lion Optimization Algorithm and the Fractional theory. The experimentation is performed using the KDD Cup 1999 dataset 1 and the HCR Lab dataset 2 and the results are evaluated based on accuracy, TPR, TNR. The metrics, accuracy, TPR, and TNR, measure the performance of the proposed Crow-AFL algorithm has shown better performance with the value of and 0.8071, 0.8813 and 0.9486 and the proposed I-AHSDT has 0.8153, 0.8903, and 0.94874, respectively newline