基于人工免疫网络模型的学习和分类算法研究  

Research on Learning and Classification Algorithms Based on Artificial Immune Network Model

作　　者：孙萧寒[1] Sun Xiaohan(College of Network Security and Information Technology,Weinan Normal University,Weinan Shaanxi 714099,China)

机构地区：[1]渭南师范学院网络安全与信息化学院,陕西渭南714099

出　　处：《信息与电脑》2018年第24期64-65,68,共3页Information & Computer

基　　金：陕西省自然科学基金(项目编号:2017JM6036);陕西省军民融合项目(项目编号:18JMR07);渭南师范学院科学研究项目(项目编号:18YKF04);电子信息(计算机技术)硕士学位点建设项目(项目编号:18TSXK06)

摘　　要：受生物免疫系统的启发,笔者将免疫网络理论应用于学习策略,提出建立一种新的人工免疫网络模型,在此基础上设计了机器学习和识别算法,提出的人要免疫网络模型既考虑了抗原和抗体之间的结合,又考虑了抗体之间的抑制作用,通过对刺激阈值和抑制阈值的学习,得到了识别算法的依据。经实践证明,提出的模型和算法能够以较高效率完成学习过程并具有较高的识别准确率。Inspired by the biological immune system,the immune network theory is applied in the learning strategies.A novel artificial immune network model is created and based on which,a machine learning algorithm and an identification algorithm are proposed.The proposed model take the combination between antigens and antibodies and the suppression among antibodies into the structure.The stimulation threshold and the suppression threshold obtained by learning are the criteria of identification.The model and algorithms proposed could learn the knowledge in a high efficiency and identify the antibodies in a high accuracy.

关 键 词：免疫网络 机器学习 识别算法 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]