Malware Analysis Using Supervised Machine Learning

Abstract

A research endeavor in the field of cyber security is being carried out under the working title of "Malware Analysis Using Supervised Machine Learning." For the purpose of identifying malicious software in the system, this initiative makes use of supervised machine learning. This endeavor relied heavily on using primary sources for its material. There is also something called dynamic malware analysis, which is when the software is analyzed in the malware analysis facility after it has been run. The environment is implemented on a Flare VM running the Windows 10 distribution. In order to acquire the log files, spyware and good ware must first be executed on the Flare virtual machine. The bag of words natural language processing methodology is used to retrieve data from these log files, and the collected data is then labelled. After this comes the process of teaching and evaluating the models. Four distinct algorithms, namely Random Forest, Decision Tree, Logistic Regression, and Support Vector Machines, were utilized in the process of training the model.