Techniques of Malware Detection: Research Review
Elshan Baghirov
Abstract
Analysis, and detection of malicious software play a crucial role in computer security. Signature-based malware detection methods were a classical solution in this area. However, malware creators are able to bypass these detection methods using some obfuscation methods like metamorphism, polymorphism. To address this issue, methods based on machine learning have been applied. However, some challenges are still present. This work presents a planned and detailed review of the malware detection mechanisms used by researchers. For this purpose, scientific works on malware detection topics were classified according to applied methods of malware detection, the accuracy of detection, etc. Several scientific works have been reviewed for analysis, and the current situation in the fight against malware has been analyzed. The main contributions of this paper are to provide detailed information to researchers about challenges on malware detection, to present to researchers a general overview of the malware detection field, to provide valuable information about tools and malware datasets that are commonly used by researchers.