A Systematic Literature Review: Usage of Logistic Regression for Malware Detection

Abstract

Malwares are serious threats since decades and now they are becoming a huge risk due to the increasing nature of their attacks. At first computer virus named “brain” was introduced, which raised the need for security measurement. Later on, the malware and malicious content did not only breach the security measurements through attaching infected devices to computer systems but also approach via network usage. Nowadays malware is a more crucial and important topic that needs to be examined carefully to avoid security issues. Millions of new malwares are reported every year so we need a fast, reliable, and trustworthy solution against this malware. Machine learning techniques are very efficient and robust to recognize malicious malware attacks. Different malware detection approaches have been developing to overcome security issues. Among all, the Logistic Regression classifier is very suitable to deal with a large number of data sets available over the internet. This article provides a step-by-step approach to conducting a Systematic Literature Review (SLR) in the domain of malware detection. SLR assesses the question of interest-based on the quality level and magnitude of existing literature. Preferred reporting item for systematic reviews and meta-analysis is used here to create a framework for SLR and verify the quality of articles. All the papers collected from various resources such as IEEE Xplore, Wiley Library, ACM Microsoft, etc. will be able to detect malware attacks.