Malware Detection Based on Machine Learning Methods, Analysis, and Tools

• Published in: Computing, Networking and Communications (ICNC), International Conference on pp. 1 - 11
• Main Authors: Almuqren, Almaha, Frikha, Mounir, Albuali, Abdullah, Amin Almaiah, Mohammed
• Format: Conference Proceeding
• Language: English
• Published: IEEE 05.11.2024
• Subjects: Feature extraction; Hazards; Machine learning; Machine learning algorithms; Malware; malware analysis and; malware detection; Open source software; Software algorithms; Systematic literature review; Systematics
• ISSN: 2473-7585
• DOI: 10.1109/ComNet64071.2024.10987343

Malware identification is essential for safeguarding digital systems from cyber attacks, and machine learning techniques are proving to be efficient in this field. This systematic literature re-view thoroughly analyzes research undertaken from 2020 to 2024 on malware detection using machine learning techniques, with a specific emphasis on analysis and tools. This research follows systematic literature review procedures to identify notable developments and trends in machine learning-based malware detection. The examination focuses on research approaches, highlighting the use of supervised, unsupervised, and deep learning algorithms. The research also examines the use of fundamental tools and frameworks for machine-learning-based malware detection. The assessment includes open-source software tools used for data preparation, feature extraction, and model evaluation. The research also explores the practical consequences of using machine learning for detecting malware. The evolution, classification, trends, and possible hazards linked to malicious software, specifically on Android and network platforms, have been the subject of an abundance of research. Nonetheless, a significant research void persists concerning the effectiveness of utilizing machine learning techniques to analyze and detect malware on di-verse platforms. By presenting a methodical examination of diverse malware analysis and detection approaches, this article attempts to bridge this disparity. The article discusses obstacles and constraints faced in machine learning-based malware detection, including dataset imbalances and adversarial assaults. The text explores possible solutions and future research paths to address these difficulties, emphasizing the importance of creating strong and adaptable detection systems. This study is a great resource for academics and practitioners that aim to strengthen cybersecurity measures and effectively address changing cyber threats.