Система динамічного аналізу шкідливого програмного забезпечення на основі інтроспекції віртуальних машин та методів машинного навчання

Abstract

Cyber wars and cyber attacks are a serious problem in the global digital environment. Technological progress is forcing malware authors to create more and more advanced and sophisticated malware. Such programs are almost impossible to detect with static analysis. Even when using dynamic analysis, a malicious file can recognize being executed by the virtual environment and change its code. Therefore, this study aims to create a dynamic analysis system, where the executable file is not able to detect being observed and can show its proposed in this article and fed to a machine learning algorithm - a support vector machine.