Effectiveness Of Detecting Android Malware Using Deep Learning Techniques

Abstract

The pervasive adoption of Android in the smartphone market has attracted the attention of malicious actors who continually exploit its open system architecture. A number of cybercriminals have been targeting Android in recent times due to its popularity. As a result of the increasing demand for smartphones, malicious users have recently been drawn to Android and taken advantage of its open system design to commit crimes. As Android has grown in popularity, attackers have been targeting it more. It is possible to gain access to data hidden from view using algorithms even though security measures have been implemented. An Android malware detection system based on machine-deep learning has been developed by utilizing dynamic analysis, in which suspected malware is executed in a secure environment in order to observe its behavior, as well as static analysis, in which malware files are examined without being executed on an Android device. As a result of our experimental results, our suggested models have a higher accuracy rate than industry standards, with a static accuracy rate of 99 and a dynamic accuracy rate of 98 for CNN-LSTM