Android Malware Detection Using Hybrid Analysis and Machine Learning Technique

This paper proposes a two-stage Android malware detection and classification mechanism based on machine learning algorithm. In this paper, we use the static analysis method to extract the software's package features, permission features, component features and triggering mechanism. Then we use the dynamic analysis tools to obtain the dynamic behavior characters of the software, and format the static and dynamic features. Finally, we use the machine learning algorithm to deal with the feature eigenvectors in two stages, and then we will get the malicious classification of the software. The experimental results show that in the data set used in this paper the proposed method based on the combination of dynamic and static malicious code detection is more accurate than the common detection engine, and the ability of classifying malicious family is much stronger.