摘要
传统光纤网络异常数据隔离算法分割类别间界限非常明确,不符合实际情况,隔离结果不可靠。为此,提出一种新的基于改进关联聚类的光纤网络异常数据隔离算法。通过建立光纤网路数据分布结构模型获取异常数据特征分布状态。通过关联规则挖掘算法获取相似度对传统聚类算法进行改进,依据相似度对隶属度进行计算。在此基础上,结合模糊集合理论提出一种改进的关联聚类算法,将该算法应用于光纤网络中。对光纤网络中不同样本进行归类处理,依据光纤网络异常数据特征分布确定异常数据类,将其隔离。实验结果表明,所提算法隔离结果可靠。
The traditional optical fiber network anomaly data separation algorithm has a very clear boundary between the classes.It does not conform to the actual situation,and the isolation results are unreliable.Therefore,a new algorithm based on improved association clustering for outlier isolation in optical fiber networks is proposed. By establishing the data distribution structure model of optical fiber network,the characteristic distribution of abnormal data is obtained.Through the association rule mining algorithm to obtain similarity,the traditional clustering algorithm is improved,and the membership is calculated according to the similarity. On this basis,combining the fuzzy set theory,an improved association clustering algorithm is proposed,which is applied to the optical fiber network.Classify the different samples in the optical network,and isolate the abnormal data class according to the characteristic data distribution of the optical fiber network,and isolate them. Experimental results show that the proposed algorithm is reliable for isolation.
作者
贡晓静
GONG Xiaojing(Guangxi Police College,Nanning 530028,China)
出处
《激光杂志》
北大核心
2018年第8期193-196,共4页
Laser Journal
基金
2018年度广西高校中青年教师基础能力提升项目(No.2018KY0713)
广西公安科学研究与技术开发计划立项项目(No.GAT2016-8)
关键词
改进
关联聚类
光纤网络
异常数据
隔离
improvement
association clustering
optical fiber network
abnormal data
isolation
