该文对入侵检测的现状进行了分析,在此基础上重点研究了数据挖掘算法在异常检测和误用检测中的具体应用。对于异常检测,主要研究了分类算法;对于误用检测,主要研究了模式比较和聚类算法,在模式比较中又以关联规则和序列规则为重点研究对象。最后对目前数据挖掘算法在入侵检测中应用所面临的难点进行了分析,并指明了今后的研究方向。 This paper analyzes the status quo of intrusion detection, and on this basis, focuses on the application of data mining algorithm in anomaly detection and misuse detection. For anomaly detection, we mainly study the classification algorithm. For misuse detection, we mainly study the pattern comparison and clustering algorithm, and focus on the association rules and sequence rules in pattern comparison. At last, the difficulties that the present data mining algorithms face in the application of intrusion detection are analyzed, and the future research directions are pointed out.