ANDROID'S DEVISED SECURITY FROM MACHINE LEARNING

Abstract

 With the exponential growth of Smartphone users and in particular Android users, are posing as vulnerable targets for cyber criminals. With adequate information available on these hand-held devices, the possibilities to exploit these are in high number. Through the years of android's development many strategies have also been put forward to detect and secure its perimeter. An alarming number of increase in Smartphone users have presented the mobile research with challenges to focus on critical threats and vulnerabilities. With Machine Learning as the medium to understand the detected malware and detection of further coming threats; In this paper, different Machine learning techniques have been reviewed which are used to develop useful models that could predict better about the malicious activities. Increasing demand for security has already overstepped static methods and signature based methods of security. Secluding vulnerabilities have led to extensive use of techniques like anomaly based detection schema, behaviour based detection and dynamic analysis of codes. The popular methods include code transformation, environment aware approaches, these entitle for malware detection.