Android malware detection using 3-level ensemble

摘要: With the dramatic increasing ofthe number of Android malware and technique avoiding detection being more sophisticated, traditional techniques based on signature is facing many difficulties. Recently, researchers are focusing incorporating machine learning algorithms. Besides effort extracting features trying new algorithms, there another way named ensemble to improve accuracy. Stacked generalization, as a powerful method, showed priority in performance by combining multiple base We proposed method that incorporates T-SNE (t-Distributed Stochastic Neighbor Embedding) algorithm into stacked generalization architecture Though devised for visualize high-dimensional data, it does provide some help classifiers when we use feature because reveals structure data at different scales. Experiments show with significantly improved