摘要
对网络环境中用户服务信息进行检测,能够有效提升网络用户服务质量。由于网络环境的复杂性和多变性,致使用户服务信息在传输过程中排列混乱。传统方法主要通过对采样的用户服务信息进行重新排序,减少用户服务信息出现排列混乱的问题,但未对重新排序后的用户服务信息进行搜索,导致检测精度低。提出基于改进半监督聚类的网络环境中用户服务信息检测方法,先利用半监督聚类方法,从全部的用户服务信息中随机抽取适当的信息数据样本,对抽取的样本数据进行标准化处理,计算频数,用标记数据确定的初始化中心点,引入BP人工网络神经,训练出最优网络权值,组建基因矩阵,计算适应值函数,对该适应度值进行搜索,以其结果为核心实现网络环境中用户服务信息检测。仿真结果证明,所提方法检测精度高,有效地提升了网络服务质量。
This paper proposes a detection method of user service information in network environment based on improved semi - supervised clustering. Firstly, the the semi - supervised clustering method is used to randomly select appropriate information data samples from all user service information, then the selected sample data are normalized and the frequency is calculated. Secondly, the initialization center is determined through labeled data, and BP artifi- cial neural network is introduced to train the optimal network weight. Moreover, the gene matrix is built and fitness function is calculated. Thus, through the search of fitness value, taking the results as the core, we complete the detection of user service information in network environment. Simulation results show that the proposed method has high detection accuracy, which effectively improves the quality of network services.
出处
《计算机仿真》
北大核心
2017年第12期309-312,共4页
Computer Simulation
基金
公益性行业(气象)科研专项重点项目(编号:GYHY201406030)
关键词
网络环境
用户服务信息
检测优化
Network environment
User service information
Detection, optimization
