Abstract: A novel abnormal detection algorithm based on artificial immune clustering is presented. The outlier measure of selection abnormal item from dataset based on distance is employed in this algorithm, which makes it easy to get better tradeoff between detection rate and false positive rate according to security policies chosen by user. Experimental results show that because this algorithm can label the abnormal activities without training dataset, it can be applied in different network system and application environments.

Key words: intrusion detection, abnormal detection, artificial immune clustering

摘要： 提出一种基于人工免疫聚类的异常检测算法，采用基于距离的异常度量因子，可以方便地筛选数据集中最突出的异常数据，能够依据不同的安全策略调节异常容忍因子，从而平衡检测率和漏报率之间的矛盾。实验结果表明，该算法采用无标记的训练数据集，能自动适应不同的网络及应用环境。

关键词: 入侵检测, 异常检测, 人工免疫聚类

CLC Number: 

• TP391