MALWARE DETECTION USING MACHINE LEARNING

Abstract

While aiming to reduce the number of negatives, we propose various functions where different machine learning can be used to get the difference between malware files and clean files. In this paper, we first use cascaded singlesided perceptrons to illustrate the idea behind our framework, and then use cascaded nucleated singlesided perceptrons. After successful testing of the average malware and clean archive, the idea behind the framework was sent to the expansion process, which allowed us to solve large malware and collect clean data.