网络入侵检测中的不确定性推理方法

Uncertain reasoning in network intrusion detection

在分析常规的入侵检测方法难以适应网络环境的问题及用户信息和专家知识的模糊性和不确定性基础上，提出将人工智能中的不确定性推理方法应用于网络入侵检测系统．给出了一种基于模糊关系的不确定性的传播方法，该方法可处理在入侵检测中遇到的模糊的、不确定的信息，从而有效地判断出用户的入侵行为．

By analyzing the problem that the known methods cannot detect the network intrusions efficiently, and studying the fuzzy and uncertain properties of both the information about the network users and the expert knowledge, this paper proposes the application of uncertain reasoning in NIDS, and provides a method for uncertainty translation based on fuzzy relations. The system will be able to deal with information that can be fuzzy and uncertain so that intrusion activities will be detected more efficiently.