摘要
当数据集发生非法入侵时,原数据属性会遭到破坏,且由于数据本身的不确定性及噪声等问题,入侵数据点位置的挖掘难度较大。为此,提出基于并行频繁模式增长算法(Frequent Pattern Growth, FP-Growth)的数据点位置智能挖掘方法。建立信息熵-主成分分析法融合算法(Entropy-Principle Compoent Analysis, E-PCA),对大数据降维。融合入侵检测和K均值聚类算法(Intrusion Detection Systems K-means clustering algorithm, IDS K-means算法)和并行FP-Growth算法,实现入侵数据的检测。利用邻居节点数据投票的方式实现入侵数据点位置智能挖掘。实验表明,所提方法检测入侵数据时误报率低于1.0%,数据点位置挖掘准确率高于98%,且能够精准实现正常数据与异常数据的聚类。以上结果均证明了所提方法具有更优的应用性能。
When the data set is illegally invaded,the original data attributes will be destroyed.Due to the uncertainty and noise of data,it is difficult to mine the location of the invaded data points.Therefore,an intelligent data point location mining method is proposed based on parallel frequent pattern growth(FP-Growth)algorithm.Firstly,Entropy-Principle Component Analysis(E-PCA)was constructed to reduce the dimension of big data.Intrusion detection systems(IDS)combined with K-means clustering algorithm and parallel FP-Growth algorithm were integrated to realize the intrusion data detection.Finally,the intelligent mining for intrusion data points was realized through the data voting of neighbor nodes.Experiment results prove that the false positive rate of the proposed method is less than 1.0%,and the accuracy rate of data point mining is more than 98%.In addition,the clustering of normal data and abnormal data can be accurately achieved.Therefore,this method has better application performance.
作者
乔阳阳
王丽娟
QIAO Yang-yang;WANG Li-juan(School of Information Engineering,Zhengzhou Technology and Business University,Zhengzhou Henan 451400,China;School of Electric Power,North China University of Water Resources and Hydropower,Henan Zhengzhou 450046,China)
出处
《计算机仿真》
北大核心
2023年第5期501-505,共5页
Computer Simulation
基金
河南省教育科学规划2022年度一般课题(2022YB0438)。
关键词
并行算法
数据点位置挖掘
入侵数据检测
Parallel FP-Growth algorithm
Data point mining
Intrusion data detection
E-PCA algorithm
Kmeans algorithm
