FICSEM: a learning method from one-case fitted in complex adaptive system

The computer immune system is a complex adaptive system (CAS) consisting of interdependent agents. The agents distinguish between self and non-self and then eliminate the non-self. In order to recognized the self in this computer immune system, this paper puts forward. the first-clustering and second-extracting method (FICSEM) to extract rules from the samples of self, which clusters those samples into subclasses and then extracts rules from the subclasses. This paper describes the details of FICSEM and our method not only recognizes self efficiently but also classifies the samples of self into subclasses. The system can judge its status by using the rules when classifying samples into a certain subclass.