Enhancing Malware Detection through Machine Learning: A Comparative Analysis of Random Forest and Naive Bayes Classification Systems

Abstract

Malware, a type of malicious software encompassing viruses, worms, Trojans, backdoors, and spyware, poses a grave threat to the confidentiality, integrity, and functionality of computer systems, given their integral role in everyday life. To combat the escalating sophistication of malware attacks, deep-learning-based Malware Detection Systems (MDSs) have emerged as indispensable components of both economic and national security. Utilizing a dataset sourced from a repository, our research focuses on classifying observations into benign and malicious software for Android devices, employing machine learning algorithms such as Random Forest and Naïve Bayes. The dataset comprises 100,000 observations with 35 features, and our evaluation metrics encompass accuracy, precision, recall, and F1-score. This study underscores the significance of MDSs in safeguarding against evolving cyber threats, utilizing cutting-edge machine learning techniques to bolster defense mechanisms.