【作者】 郭晨；

【导师】 梁家荣；

【作者基本信息】 广西大学 ， 计算机应用技术， 2005， 硕士

【摘要】 近年来,生物免疫系统已经成为一个新兴的生物信息研究课题。它是除神经系统外,机体能特异地识别“自己/非己”(SNS)刺激,对之作出精确应答,并保留记忆的功能系统。免疫模式SNS之后,整个免疫学界普遍流传着另外一个新的理论,那就是“危险模式”理论(danger model)。这个理论的主要创导者就是Matzinger,他认为免疫系统并不能区分自我/非我(SNS)而只区分危险信号(danger signal)的有无。这就打破了以往耐受的认识,并对传统免疫系统只区分SNS的现代免疫学基本理论提出质疑,同时“危险模式”的免疫理论可对移植、肿瘤、自身免疫反应等诸多问题做出较好的解释,目前已引起各界学者广泛关注并产生了较大影响。 危险模式理论并不是免疫应答的一种扩展方式,而是一种全新的理论模式。虽然迄今为止还没有得到生物免疫界的普遍认可,但对于对生物免疫系统的一种仿真系统人工免疫完全可以摆脱生物免疫认知上的束缚,而扩充到人工免疫领域。 基于免疫系统及危险模式理论的基础上,本文讨论了数据挖掘理论的框架体系,借鉴生命免疫科学中危险模式的免疫模型理论,围绕基于危险模式的免疫算法、关联规则挖掘和数据聚类分析等数据挖掘技术,提出了基于危险模式免疫算法的数据挖掘方法及其应用研究。更多还原

【Abstract】 Lately, immune system has been a novel subject of biologic information. It is another special system which can identify self and non-self(SNS) and response exactly .Then save the functional system. After the emergence of SNS model there is another new model go around in immunology. That is the danger model. Matzinger who is the main introducer of this new theory believed that immune system can not to identify self and non-self but to distinguish whether have the danger signal. So danger model question the traditional immune model (SNS). Using danger theory many aspects of medicinal phenomena can explain perfectly. Nowadays,this novel theory arouse many experts’interesting.The theory of danger model is not the extended methods of immune response but a novel theory model. Since now though many immunologists did not authorize this new theory. But at the field of artificial immune system we can extend this kind of model in not take the shackle of immune system into account.Based on these developments, many good algorithms for various data mining tasks are obtained.Based on previous research, the major research works in this article are:1 search of the theory of danger model inspired algorithms has been proposed. And try to create a novel a model of danger model inspired algorithms then prove the feasibility of the algorithms through out experiments.2 Then introduce this new algorithms to the areas of data mining .The major applications in data mining involve Association Rule Mining and data cluster analysis.3 In the later time the trial research of safety aspects based on danger model inspired algorithms has been proposed. Then in the paper we bring forward a adaptive mailbox and intrusion detection system based on danger model inspired algorithms.4 Finally, this paper try to set up a web-safety system using danger model inspired algorithms.更多还原