Abstract:
Android became one of the most widely used mobile operating system, and
the amount of malware targeting it is increasing at an alarming rate. Despite
the fact that notable studies on malware detection and classification have
been conducted in academia and industry, but a robust and efficient solution
for detection of all types of Android malwares is still a challenge. Existing
solutions do not adequately consider factors like concept drift and are often
not based on a hybrid approach. Also they have been designed using infor mation collected by running malware samples on virtual environment (and
not on a real device). Thus, they are not able to detect sophisticated or new
malwares. In this research work we have studied existing solutions and after
finding their limitations we have proposed an effective and efficient hybrid
Android malware detection solution based on machine learning to detect and
categorize existing, emerging and behaviour evolving Android malwares.
