Safeguarding User Privacy: Machine Learning Strategies for Android Malware Detection

Google Android as the leading smartphone operating system, boosting a significant market share and billions of active users, has rendered it receptive to a high number of malware attacks. Android malware becomes crucial as the number of malicious applications continues to rise in effective methods for detecting and classifying. In this paper, an ensemble-based machine learning approach is used for detection of Android malware, harnessing features like API calls, requested permissions, and network activity to construct a model which we can easily discern between legitimate and malicious applications. By supervising a performance evaluation through twofold cross-validation and metrics like F1-score, precision, and recall, we illustrate the efficacy of the proposed approach. The result of our study indicates that ensemble techniques, especially random forest, exhibit the highest accuracy in detecting malware applications. This underscores their significance as a valuable tool to enhance mobile security and safeguarding user privacy amidst evolving threats.