December 2022
Intermediate to advanced
190 pages
4h 56m
English
In the previous chapters, we discussed about the effectiveness of system call graph-based malware detection mechanisms. These mechanisms considered the vertex level features such as centralities for malware detection. In the current settings, it is very difficult to incorporate high dimensional edge level features such as adjacency matrix in a machine learning classifier. In order to overcome this limitation, a graph signal processing based approach was proposed in [189]. In this chapter, we first discuss how to generate various graph signals from system call sequence of an application for constructing low dimensional feature vectors. Later we discuss about employing various ...