Android Malware Detection Methods Based on the Combination of Clustering and Classification

摘要: With the popularity of Android platform, malware detection is a challenging practical problem that needs to be resolved urgently. In this paper, we propose two static analysis methods for based on combination clustering and classification. First, obtain original feature set from manifest file disassembled code applications. Then, through category appearance frequency each feature, extract some key features so as reduce dimensionality vector. Finally, classification distinguish malicious benign One mixed clustering, which clusters samples together; other separate separately. We choose use K-mean algorithm K-Nearest Neighbor (KNN) algorithm. Evaluation results show our outperform common SVM-based method in accuracy, KNN-based prediction time. addition, ability unknown families also better than method.