Optimizing Machine Learning based Large Scale Android Malwares Detection by Feature Disposal

摘要: As a favorable opening platform for mobile terminals, android attracts close attentions from large number of hackers. The great potential security hazard makes the requirement malicious software detection become effective, rapid and multitudinous. In recent years, lot machine learning based methods have been proposed. However, most focuses are searching more effective feature information. this paper, we propose an optimization method machine-learning-based malware by focusing on disposal We extract permission intent information malwares, dispose them in series effectively methods. After disposal, use several algorithms to verify their effectiveness, conclude comparing list. comparing, algorithm combining processing. effectiveness our proposal is illustrated justified extensive experimental results.