Information security professionals are doing their best to come up with novel techniques to detect malware and malicious software. One of the trending techniques is using the power of machine learning algorithms to detect malware. On the other hand, attackers and cyber criminals are also coming up with new approaches to bypass next-generation systems. In the previous chapter, we looked at how to attack machine learning models and how to bypass intrusion detection systems.
Malware developers use many techniques to bypass machine learning malware detectors. Previously, we explored an approach to build malware classifiers by training the system with grayscale image vectors. In a demonstration done by the Search And RetrieVAl of Malware (SARVAM) research unit, at the Vision Research Lab, UCSB, the researchers illustrated that, by changing a few bytes, a model...