Detection of advanced malware through machine learning

the volume, complexity, and variety of malware are continually increasing. Due to this nearly exponential growth, any prospective deployed defense has to Automatically adjust to newly emerglng patterns. Leveraging machine leaming techniques for protectron against malware allows automating knowledge extraction and this way to cope with the ever-changing threat landscape. However, accurate detection and prediction of earlier unseen malware variants remain an ongoing challenge. That becomes especially complicated with the existence of advanced malware - persistent. stealthy. multi-stage attacks, which target a specific computer system and are initially designed for continuous evasion of detection. This Ph.D. project will investigate the applicability and efficiency of the most powerful machine learning methods in order to accurately and rapidly detect and predict advanced malware in a systematic way.