A Survey on Visualization-Based Malware Detection

Abstract

In computer security, the number of malware threats is increasing and causing damage to systems for individuals or organizations, necessitating a new detection technique capable of detecting a new variant of malware more efficiently than traditional anti-malware methods. Traditional antimalware software cannot detect new malware variants, and conventional techniques such as static analysis, dynamic analysis, and hybrid analysis are time-consuming and rely on domain experts. Visualization-based malware detection has recently gained popularity due to its accuracy, independence from domain experts, and faster detection time. Visualization-based malware detection uses the image representation of the malware binary and applies image processing techniques to the image. This paper aims to provide readers with a comprehensive understanding of malware detection and focuses on visualization-based malware detection.

RFSA: A Ransomware Feature Selection Algorithm for Multivariate Analysis of Malware Behavior in Cryptocurrency

“…This provides a more holistic insight into the behavior of ransomware. The importance of utilizing AI methods, particularly ML and deep learning, for detecting and preventing the spread of malware threats was emphasized by [25]. While this approach acknowledges the significance of analysis processes in identifying malware patterns, the RFSA contributes by integrating financial aspects into the evaluation of ransomware.…”

“…While this approach acknowledges the significance of analysis processes in identifying malware patterns, the RFSA contributes by integrating financial aspects into the evaluation of ransomware. The primary focus of [25] is on the broader context of malware types, binary executables, analysis methods, and AI applications.…”

Application and optimization of multidimensional evaluation functions for faster malware detection

“…ontemporary malware detection methodologies are grounded on various metrics. However, considering the incessantly evolving nature of malware, there's a growing demand for reducing malware detection time in order to minimize system complexities and cyberattack damages [1]. Given that these metrics alone are insufficient, a novel approach is deemed necessary.This research proposes a new malware detection method combining multidimensional evaluation functions and the exploitation of energy harvesting technologies.…”