SESSION

The application of machine learning (ML) in malware detection and response has been at the forefront of cybersecurity research for decades. More and more research papers in this field achieve results that effectively push research and application forward with most malware detection engines leveraging machine learning algorithms. In particular, many recent studies have used deep learning models to learn the semantics of program execution. However, most methods have not been uniformly validated with few people having full access to thoroughly analyze their impact and effectiveness. In this presentation, we will share our methodology and results from evaluating various ML algorithms and their effectiveness against real-world obfuscation techniques used by attackers in the wild.