A Multi-Objective Hyper-Heuristic Improved Configuration of Svm Based on Particle Swarm Optimization for Big Data Cyber Security

Big Data Cyber security analytics is a growingly significant field of study and application that uses big data techniques and technology to analyses security event data in order to safeguard networks, computers, and data from illegal access. There have been several reports on Big Data Cybersecurity Analytic Systems in the literature, but an exhaustive and systematic analysis of the research from an architectural viewpoint is lacking. When In this paper, the SVM design process is organized as a bi-objective optimization problem with accuracy and model complexity as two competing goals. The system presents a novel, problem-independent hyper-heuristic framework for bi-objective optimization. It is novel to create a hyper heuristic for this situation. The hyper-heuristic structure proposed comprises of low-level heuristics and a significant level system. In light of the pursuit execution, the significant level system figures out which low-level heuristic ought to be utilized to create another SVM arrangement. Every one of the low-level heuristics look through the SVM design search space really utilizing an alternate arrangement of rules. To estimate the Pareto set of SVM arrangements, the suggested design adaptively combines the benefits of degradation with Pareto-based strategies for dealing with bi-objective enhancement. Microsoft's anomalous intrusion detection and big data malware classification have been used to gauge the efficiency of the suggested methodology.