Malware Analysis with Machine Learning Tools

Description

The book discusses the use of some neural networks in the aspect of working with malware. In particular, it discusses the use of neural networks for obfuscation and detection of malware obfuscated using neural networks. Separately, measures to improve the stability of the Snort intrusion detection system with a plug-in machine learning module during an attack on the intrusion detection system itself and the detection of polymorphic malware are considered.
As a mathematical apparatus, methods based on the k nearest neighbors, mean shift and Fadeev-Leverrier method, solutions based on «unreachable code» were used. Generative-adversarial, convolutional, recurrent neural networks with various parameters and hyperparameters were used.
The book is designed for researchers of network and system Infrastructure security when an attacker uses machine learning tools.