Abstract: Negative selection algorithm simply takes each self point and its vicinity as the self region.In this paper，real-valued negative selection algorithm is studied，the continuous self region is defined，the self data is classified to the self region through the means of cluster analysis，the partial training takes place at training stage according to the radius of self region and cosine distance with gravity of the self region，and in the self region V-detectors are deployed.Experimental results show that when the train self points are used together as a whole more information is provided，the boundary of self region can be probed，also efficiency of system and detection rate can be increased.

Key words: artificial immune system, real-valued negative selection algorithms, cluster analysis, self region

摘要： 否定选择算法将单个自体点和其邻近点作为自体区域训练检测器。研究了实值否定算法，定义了连续的自体区域，采用动态聚类法将自体样本点分类到自体区域，训练时根据自体区域半径和与自体区域重心之间的余弦距离做局部训练，并在自体区域内使用可变阈值检测器。实验证明当耐受自体点被当成一个整体使用时能提供更多的信息，可以探测出自体区域边界，使系统效率和检测率得到提高。

关键词: 人工免疫, 实值否定选择算法, 聚类分析, 自体区域