An Ensemble Learning Approach to Detect Malwares Based on Static Information

The proliferation of malware and its variants have brought great challenges to malware detection. The traditional static analysis methods are complicated and consume a lot of human resource. Moreover, most of the current detection methods mainly focus on the single characteristic of malware. To address the above issues, this paper proposes an Ensemble Learning approach method to detect malwares based on static information. The image feature and entropy features are used separately to train two models. Besides, with the guidance of ensemble learning principle, the two models are combined and obtain better accuracy compared with each of two models. We conduct comprehensive experiments to evaluate the performance of our approach, the results show the effectiveness and efficiency.