MalDAE: Detecting and explaining malware based on correlation and fusion of static and dynamic characteristics

Weijie Han, Jingfeng Xue, Yong Wang, Lu Huang, Zixiao Kong, Limin Mao

Abstract: Highlights•Explore the difference and relation between the dynamic and static API call sequences by defining a number of types of malicious behaviors.•Build a malware detection framework called MalDAE based on correlation and fusion of dynamic and static API call sequences.•Provide an explainable theoretical framework for malware detection based on fusion of dynamic and static API call information.•Conduct comprehensive experiments and a detailed comparison with similar studies to evaluate MalDAE.