Analysis of machine learning solutions to detect malware in android

The recent use of mobile devices and increase in connectivity technologies(GSM, GPRS, Bluetooth & WiFi enable us to access abundant services. These services and communication channels are exploited by susceptibilities immensely. Hence, for malware writers, mobile devices became ideal target. Applications installed on smartphones request access to the sensitive information which may lead to security vulnerabilities. Different malwares named as Botnet, Backdoor, Rootkits, Virus, Worms, and Trojans can attack android Operating System (OS). Due to these attacks privacy of the users is compromised. This paper surveys the already proposed security solutions by using machine learning approaches especially focused on supervised, semi supervised and unsupervised approaches. We also analyzed the architecture of these approaches and present the taxonomy of Android OS based security solutions. Our aim is to provide the best approach for malware detection in Android OS.