Malware Detection Using Binary Visualization and Neural Networks

Abstract

Any programme or code that is damaging to our systems or networks is known as Malware or malicious software. Malware attempts to infiltrate, damage, or destroy our gadgets such as computers, networks, tablets, and so on. Malware may also grant partial or total control over the affected systems. Malware is often detected using classic approaches such as static programme analysis or dynamic execution analysis. The exponential rise of malware variations requires us to look beyond the obvious in order to identify them before they do harm or take control of our systems. To address these drawbacks, malware detection based on binary visualisation followed by the deployment of powerful machine learning techniques such as Convolutional Neural Networks (CNN) performs better than the ones we now use. We use these discoveries in our efforts to identify malware in different files and websites. We strive to complete the objective by employing representations of malware software binaries. With this concept, we can construct a better bridge for developing a functioning model that can identify malware in real time.