A BI-OBJECTIVE HYPER HEURISTIC SUPPORT VECTOR MACHINES FOR BIG DATA CYBER SECURITY

Abstract

With the rapid growth of big data and digital technologies, cybersecurity has become a critical concern due to the increasing volume, variety, and velocity of data generated across networks. Traditional security mechanisms struggle to efficiently detect and prevent cyber threats in such large-scale environments. This project proposes a bi-objective hyper-heuristic framework integrated with Support Vector Machines (SVM) to enhance cyber security in big data systems. The approach focuses on optimizing two key objectives: maximizing detection accuracy and minimizing computational complexity. The proposed system utilizes hyper-heuristic techniques to automatically select and adapt suitable heuristics for feature selection and model optimization. These heuristics guide the learning process of the SVM model, enabling it to efficiently handle large and complex datasets. The system processes network traffic and cybersecurity data to extract relevant features, which are then used to train the SVM classifier for identifying normal and malicious activities. By incorporating a bi-objective optimization strategy, the model balances performance and efficiency, making it suitable for real-time applications. Experimental results demonstrate that the proposed approach outperforms traditional SVM and single-objective models in terms of accuracy, scalability, and computational efficiency. The hyper-heuristic framework enables adaptive learning, allowing the system to handle dynamic and evolving cyber threats effectively. However, challenges such as parameter tuning and high computational requirements for large datasets remain. Overall, this project highlights the potential of combining hyper-heuristic optimization with machine learning techniques to develop intelligent and scalable cybersecurity solutions.