A Review on Behavior-Based Detection for Network Threats

With the appearance and development of the technology of malicious codes and other unknown threats, information security has drawn people's attention. In this paper, we investigate on behavior-based detection which is different from traditional static detection technology. Firstly, we discuss the procedure in detail, especially feature extraction and classification. Several machine learning methods such as SVM, bayesian decision are adopted to solve these problems. What's more, we outline the architecture to specify operating mechanism. The features and conclusion are discussed finally.