PE File-Based Malware Detection Using Machine Learning

Abstract

In current times, malware writers write more progressive sophisticatedly designed malware in order to target the user. Therefore, one of the most cumbersome tasks for the cyber industry is to deal with this ever-increasing number of progressive malware. Traditional security solutions such as anti-viruses and anti-malware fail to detect these advanced types of malware because the majority of this malware are refined versions of their predecessor. Moreover, these solutions consume lots of computational resources on the host to accomplish their operations. Further, malware evades these security solutions by using intelligent approaches such as code encryption, obfuscation and polymorphism. Therefore, to provide alternatives to these solutions, this paper discusses the existing malware analysis and detection techniques in a comprehensive/holistic manner.