نوع مقاله : مقاله پژوهشی
عنوان مقاله English
نویسنده English
This paper presents a novel hybrid framework for Android malware detection that integrates an Improved Particle Swarm Optimization (IPSO) algorithm with a Fuzzy Kernel Support Vector Machine (FKSVM). The primary objective is to overcome the challenges of traditional methods, including high feature dimensionality, premature convergence, and the inability to model the inherent uncertainty in malware data. The IPSO algorithm, leveraging a dynamic inertia weight and a random mutation operator, prevents premature convergence and selects an optimal feature subset. Subsequently, the FKSVM model effectively manages data uncertainty through a novel fuzzy kernel. Evaluation of the proposed framework on the CIC-AndMal2020 dataset demonstrates that the method achieves a binary classification accuracy of 99.98% and an extremely low false alarm rate of 0.01%. These results clearly indicate the superior performance of the proposed framework compared to existing methods in terms of accuracy, speed, and robustness against adversarial attacks, thereby confirming its effectiveness.
کلیدواژهها English
