Empowering Digital Resilience: SVM Kernel Comparison for Malware Classification to Enhance Security

Malware is a software design built specifically to infiltrate the digital system, breach the security of the device and have increased potential to cause serious damage to individual and organizational data. As a result, it is crucial to integrate the ML technological capabilities to detect the malware and categorize them into benign and suspicious. The ML-supervised classification approach utilizes the labels linked to the various parameters to categorize the application as Benign and Suspicious. For the categorical analysis, the SVM algorithm is used and the comparison of the three different SVM kernels: Linear, Polynomial and Radial Basis Function (RBF) kernel is done based on performance evaluation criteria. The best-performing SVM kernel is RBF with an accuracy of 98%. To make a sustainable digital space it is necessary to take necessary steps to safeguard the risk of malware applications.