Malware Classification using Malware Visualization and Deep Learning

Abstract

The presence of malware and potential attack has posed a threat to cyber security. The potential challenges in malware detection is that the increasing number and variety of unknown malware makes it impossible to identify its existence. This research study has proposed a novel method for categorizing malware executables based on their visual representation by converting the malware binaries to grayscale images and then classifying them using CNN. The main objective of this research work is to employ several models, which will then be used to perform a comparison study on various outcomes to demonstrate the applicability of utilizing the described approaches to visually categorize the malware.