Android Malware Category and Family Classification Using Static Analysis
Cong‐Danh Nguyen, Nghi Hoang Khoa, Khoa Nguyen-Dang Doan, Nguyen Tan Cam
Abstract
In recent years, Android malware has been overgrown, challenging malware analysts. However, there has been a lot of research in detecting and classifying Android malware based on machine learning. Android malware classification is an essential goal in classifying malware families. This paper proposes the application of machine learning and deep learning methods in classifying malware families and categories based on many different datasets to evaluate and select suitable methods for each dataset. This work demonstrates that with the Drebin and CICMaldroid2020 datasets classified by family and category, respectively, after feature extraction and selection, trained and evaluated with machine learning models, results are high accuracy, and the false positive rate is low. We also compare our results with several previous studies to highlight our results.