Klasifikace malware za použití ANN a příbuzných metod

Abstract

Nowadays, the number of computer threads is rising rapidly. This is making manual analysis and signature-based malware detection obsolete. Due to increase in computational power and improvements in the machine learning field, the usage of neural networks for analysis seems to be a very promising alternative. The purpose of theoretical section of this thesis is to define neccessary terms, principles and definitions in order to understand following practical part, which goal is to implement and describe malware classification based on malware families.