Deep Generative Model for Malware Detection
Yitu Fu, Qing Lan · 2020
Malware detection play different roles in a computer system and exhibit high degrees of importance with respect to system security. Malware detection is the process of attempting to infer the reputation score of the files via the applications. However, malware detection approaches are challenged by the large, dynamic, and heterogeneous space of benign binaries that they must track. In this research, we use deep generative models to develop two a semi-supervised Bayesian models for malware detection, in which we model the data generating process to be dependent on a Gaussian mixture. Furthermore, we propose the efficient stochastic gradient optimization technique used in deep generative models makes our model suitable for large data sets. Extensive experimental results on one real-world dataset demonstrate that our model is the effectiveness. Moreover, the semi-supervised deep generative scheme achieves comparable or even better results in malware detection when compared with classic and alternative machine learning models. This demonstrates the feasibility of deep generative model and presents a promising new approach to malware detection.