High Accuracy Android Malware Detection Using Ensemble Learning

摘要: With over 50 billion downloads and more than 1.3 million apps in Google's official market, Android has continued to gain popularity among smartphone users worldwide. At the same time there been a rise malware targeting platform, with recent strains employing highly sophisticated detection avoidance techniques. As traditional signature-based methods become less potent detecting unknown malware, alternatives are needed for timely zero-day discovery. Thus, this study proposes an approach that utilises ensemble learning detection. It combines advantages of static analysis efficiency performance machine improve accuracy. The models built using large repository samples benign from leading antivirus vendor. Experimental results presented shows proposed method which uses feature space leverage power is capable 97.3-99% accuracy very low false positive rates.