Mobile Botnet Detection: A Machine Learning Approach using SVM

Abstract

Abstract: Android is now the most widespread mobile operating system worldwide. Over the years the volume of malware targeting Android has continued to grow. This is because it is easier and more profitable for malware authors to target an operating sys-tem that is open-source, more prevalent, and does not restrict the installation of appsfrom any possible source. Hence, in this paper we present a deep learning approach that leverages Support Vector Machine (SVM) forAndroid botnet detection. The SVM model employs 342 static features to classify new or previously unseen apps as either ‘botnet’ or ‘normal