With the emergence of deep learning, recent years have witnessed a booming of artificial intelligence (AI) applications and services in many fields of modern society, ranging from face recognition to video surveillance to many recommendation systems. In addition, there is a tremendous increase in cyberattacks or data leakage which draws much attention from both the professional and the public due to the popularity of mobile devices like Android. Driving by this trend, many security approaches push AI or deep learning to Android malware detection to enhance the accuracy and capability of scanning or detecting systems. However, research on this field is still in the early stage since there is the diversity of deep learning models and a shortage of standardized Android datasets for evaluating the efficiency of each approach. To this end, we conduct a comparative study of famous deep learning models on one dataset to give an overview of their powers in providing the method of dissecting Android malware.
