Classification Method for Malware Detection on Android Devices

Due to its popularity and open-source nature, Android is the mobile platform that has been targeted the most by malware. Android allows downloading and installation of apps from other unofficial market places. This aims to steal personal information or to control the users’ devices. More specifically, malware attacks private and financial information on mobile payment applications and networks, and thus is very sensitive. In this paper, we propose an efficient malware detection model for Android devices centered on mobile payment applications. This model is based on client/server architecture to reduce the heavy computations of data on the mobile device and doing the processing remotely on the server. Our approach aims to develop an optimized algorithm based on machine learning models to extract the permissions and for better classification of the new installed applications on Android devices. The Random Forest regression algorithm with the numerical ranging from −100 (benign) to 100 (malware) gives good results and an accuracy close to 100%. Therefore, the proposed model is suitable to secure the Android devices in the mobile commerce context.