Malware detection through binary analysis

Abstract

Malware Detection is an important problem in modern day due to the increasing frequency of malware attacks using unknown malware strains. Unlike traditional detection techniques which require a signature for each sample, binary analysis relies on the structure of the program as well as features corresponding to the binary to determine whether it is a malware or not. In this work, we are using static features from various malware samples and use machine learning models to determine whether a given sample corresponds to the presence of a malware or not. In order to have this working in real time, we only use features obtained from the binary file and its corresponding assembly file which can be generated from the binary