%0 Journal Article
%T An Adaptable Software Architecture for Artificial Immune System<br>一个适应性的人工免疫系统的软件构架
%A YAN Han
%A ZHANG Kun
%A LI Qian-Mu
%A LIU Feng-Yu
%A <br>严悍
%A 张琨
%A 李千目
%A 刘凤玉
%J 计算机科学
%D 2006
%I 
%X Adaptability is one of the most significant properties of AIS (Artificial Immune System). However, AIS software development and application are facing two complex and tangled adaptability problems: the evolution of data sources and evolution of multiple learning algorisms. We expand and improve on the existing AIS architecture, and propose an adaptable software architecture for AIS. The architecture is gene computing-centered, with meta-gene expanded to adapt to the evolution of data sources, and with pluggble learning algorism components and algorisms validation mechanism to adapt to the multiple algorisms revoution. With this architecture, evolution of data sources will be independent with learning algorisms; on the other hand, learning algorisms reuse for multiple data sources and evolve independently. This architecturesimplifies complexity of AIS software, and hence improves the efficiency of AIS development and application. Further, it is helpful to realize future self-adaptive immune computing.
%K Adaptability
%K Artificial immune system
%K Software architecture
%K Meta-gene
%K Evolution computing<br>适应性
%K 人工免疫系统
%K 软件构架
%K 元基因
%K 进化计算
%U http://www.alljournals.cn/get_abstract_url.aspx?pcid=5B3AB970F71A803DEACDC0559115BFCF0A068CD97DD29835&cid=8240383F08CE46C8B05036380D75B607&jid=64A12D73428C8B8DBFB978D04DFEB3C1&aid=5BB5914E5581286D&yid=37904DC365DD7266&vid=27746BCEEE58E9DC&iid=5D311CA918CA9A03&sid=96A53C367B5173D7&eid=4FE459D71E3BF8EB&journal_id=1002-137X&journal_name=计算机科学&referenced_num=0&reference_num=8