Static Analysis of Android Malware Detection using Deep Learning

Abstract

Android Malware is very common these days as applications are not created by trusted sources. People enter their personal data, save cards and much more, thinking these apps are going to keep them fit or help remind them to do certain essential works which we tend to forget in this busy routine of life. In such cases, detecting the malware before even installing an application would be of great help to us. It could possibly even stop a few crimes. In this paper, we propose to use the fully connected deep learning model for detection of Android malware. Key features of the proposed work include detection of Android malware even before installation, the name of the Android malware, version packages with proven extremely high accuracy of about 94.65%. This model also learns all features from all combinations of features. It includes extensive research and testing to achieve very high accuracy.