Using Your Head: Identifying Windows Malware by Deep Learning on PE Headers

An analysis into utilizing deep learning NN models like multi-layer perceptrons and fully-connected nets to learn if certain features in Windows PE header files are indicative of malware or not.

Technologies used: deep-learning models (multi-layer perceptrons (MLPs), fully-connected networks (FCnets)), supervised learning algorithms (random forest), dataset collection, malware analysis, Windows PE headers