大数据网络中用户特征数据准确检测仿真

摘要

This article proposes a method for accurate detection on data of user characteristic in large data network based on estimation of density center of mean value.Firstly,the author used feature extraction method to carry out preliminary classification and feature recognition for the characteristic data and calculated degree of association among each user characteristic integrated with detection principle of degree of association.Then,on that basis,the method of center estimation of mean value density was used to acquire cluster center of the characteristic data.Moreover,the research worked out set of center density of the user characteristic taking all user data points as kernel and set judging threshold value of data type of user to screen data of user characteristic.Finally,the accurate detection was completed.Simulation results show that the method can achieve the detection of data of user characteristic preferably.It has better application value.%对大数据网络中用户特征数据进行准确检测,可对用户进行准确定位,提高用户的服务质量.进行特征数据检测时,应对大数据网络中用户特征数据间的关联度进行计算,获取数据聚类中心,并结合数据类型判断阈值进行用户特征数据数据筛选完成检测,但是传统方法通过将用户特征数据周期定义为一个函数,依据不同特征的支持度和置信度完成检测,但是无法利用数据的关联性获取聚类中心,不能计算数据类型判断阈值,降低了检测的准确性,提出一种基于均值密度中心估计的大数据网络中用户特征数据准确检测方法.先利用特征提取方法对用户特征数据进行初步的分类与特征识别,融合于关联度检测原理对各用户特征间的关联度进行计算,在此基础上采用均值密度的中心估计方法获取用户特征数据数据聚类中心,计算出以全部用户数据点为核心的用户数据中心密度集合,设定用户数据类型判断阈值进行用户特征数据数据筛选,完成了大数据网络中用户特征数据准确检测.仿真结果表明,所提方法可以较好的实现对大数据网络中用户特征数据数据的检测,而且具有很好的应用价值.