Statistical Analysis Of Malware In Anroid With The Techniques Of Machine Learning

  • Hemant Kumar, Akshay Chamoli, Subodh Kuma


 The use of smartphones continues to sour, with Android leading the way. Google claims to have around 1.4 billion active mobile devices across the globe. According to The Telecom Regulatory Authority of India, India has become the second country globally with a user threshold of 1.03 billion users. The massive user base has caught the eyes of cybercriminals trying to attack Android using malware. Malware is malicious software created to destroy computer or electronic systems without the knowledge of the user using the system. This work aim has created an efficient malware detection system with machine learning-enabled features for Android Environment so that they can sense malware applications with the help of static features. The system extracts various permissions and API Tags from the Android applications. It uses these features along with five unique machine learning techniques for classifying whether the application is malicious or benign. The experimental results are promising as we have achieved high accuracy in detecting Android malware with all the classifiers. We have also analyzed the effect of a number of Android applications used in the database on the accuracy of classifiers.