Visualization and Deep Learning Approaches for Malware Analysis

Tóm tắt

Malware is one of the most serious threats in the field of cybersecurity today. Malware analysis plays an essential role in detecting, preventing, and building effective defense solutions. This paper reviews traditional malware analysis methods, including static and dynamic analysis, and introduces a visualization-based approach combined with Convolutional Neural Networks (CNN). Experimental results show that the proposed method achieves high accuracy (95%), which indicates its potential as an effective solution in the context of increasingly sophisticated malware variants.