大数据下网络化软件时间变化敏感点检测仿真  

Networked Software Time Change Sensitive Point Detection Simulation under Big Data

作　　者：刘红英 LIU Hong-ying(Guangzhou College of Technology and Business,Guangdong Guangzhou 510850,China)

机构地区：[1]广州工商学院

出　　处：《计算机仿真》2019年第12期353-356,395,共5页Computer Simulation

摘　　要：网络化软件时间变化敏感点变化较为频繁,当前的方法无法保证检测的正确率,而且耗费时间较长。为了能够有效的对敏感点实施检测,引入大数据的对比方法,提出大数据下网络化软件时间变化敏感点检测方法。方法先利用高斯核函数对阈值进行确认,通过对大数据模糊聚类方法对敏感点数据进行分类,并获取出它的特征向量,对获取的敏感点数据特征集进行大数据降维处理,接着利用非线性函数把获取的敏感点数据信息映射到高维特征空间中,在特征空间中确定软超球体对敏感点进行判定检测。仿真证明,所提方法在对网络化软件时间变化敏感点检测时,能够在较短的时间内完成正确率较高的检测。This article puts forward a method to detect the sensitive points of time change of networked software based on big data by introducing the comparison method of big data.Firstly,this method used Gaussian kernel function to confirm the threshold value.Secondly,the data of sensitive point was classified by the fuzzy clustering method based on big data,so that its feature vector was obtained.Thirdly,we performed the big data dimensionality reduction on the data feature set of sensitive point,and then we used the nonlinear function to map sensitive point data information into the high-dimensional feature space.Finally,we determined soft hypersphere in feature space and thus to decide the sensitive points.Simulation results show that the proposed method can complete the detection with higher accuracy in a shorter time.

关 键 词：大数据 网络化软件 敏感点检测 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]