Android Malware Detection Using Machine Learning
جاري التحميل...
التاريخ
المؤلفين
عنوان الدورية
ردمد الدورية
عنوان المجلد
الناشر
IOSR Journal of Computer Engineering
خلاصة
As the popularity and ubiquity of Android devices continue to rise, so does the risk of malicious software targeting these platforms. Android malware poses significant threats to users’ privacy, data security, and overall device performance. Therefore, effective detection and mitigation of Android malware have become essential to ensure a safe and secure user experience. In this project, a malware detection system is proposed that extracts permission and intent features from APK files using the SISIK web tool to effectively identify and classify applications as malware or benign without the need to run the application. This is done by incorporating two different Machine Learning (ML) algorithms, which are Random Forest (RF), and Support Vector Machine (SVM). To obtain the best performance in our system, we use a feature selection method. The main contribution of this Research Paper is to enhance the security of Android devices detecting malicious applications before installing them in the devices. Our results show that the RF model, with the use of the Genetic algorithm (GA) to reduce the dataset’s dimensions, achieved the highest performance metrics, including accuracy, recall, F1 score, and precision of 98%, 99%, 98%, and 98%, respectively.