基于阴性选择算法的改进模型  

The improved model based on negative selection algorithm

作　　者：傅龙天[1] 陈腾林[2] 

机构地区：[1]福州外语外贸学院,福建福州350011 [2]闽江学院,福建福州350011

出　　处：《黑龙江工程学院学报》2016年第5期50-53,共4页Journal of Heilongjiang Institute of Technology

摘　　要：人工免疫学中的阴性选择算法是其核心算法,在各行业应用广泛。但其不足之处也越来越明显,例如在训练样本选择方面、训练学习算法方面,都有可能影响检测精度。训练学习算法中引入半监督学习机制,并在样本选择上扩展训练样本来源,使训练学习更有针对性。仿真实验证明改进后的模型能提高检测率,并具备较强的自适应能力。Artificial immunology loyalty is the core algorithm of negative selection algorithm ,which has been widely applied to various industries .But its deficiency is becoming more and more noticeable ,such as the training sample selection and learning algorithm are likely to influence the accuracy of detection .Learning algorithm is introduced in this paper as a semi‐supervised learning mechanism ,of which the training sample source in the sample selection is expanded to make the training more targeted .Simulation experiment proves that the improved model can improve the detection rate ,and have strong adaptive capacity .

关 键 词：阴性选择算法 人工免疫 半监督学习 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]