大数据环境下网络异常数据准确检测仿真

摘要

大数据环境下对网络异常数据的准确检测能够确保网络系统的安全性.传统方法进行异常数据检测时,需要对全部数据的不同样本进行不同程度的扰动,过程较为复杂,且无法彻底抑制大规模冗余数据干扰,存在数据检测正确率低的问题.提出基于改进关联规则的大规模冗余数据干扰下的网络异常数据准确检测方法.定义异常数据的支持度和置信度,并给出最小支持度阈值和最小置信度阈值,利用混沌理论从网络全部数据序列中获取描述异常数据特征的混沌数据特征关联,定义正常数据间的紧密性和异常数据间的紧密度量,并得到当前各聚类内数据点的紧密性程度大小,利用数据特征聚类的紧密性与分离性方法对含有大规模冗余数据干扰的网络环境进行抑制,完成对网络异常数据准确检测.仿真结果表明,所提方法检测精确度高,可有效地保障网络安全稳定的运行.%An accurate detection method of network abnormal data under the large-scale redundant data interference based on the modified association rule is proposed.Firstly,the support degree and confidence degree of abnormal data are defined,and the minimum threshold value of support and confidence degree is given out.Then,the character association of chaos data describing the abnormal data feature from all network data sequence is acquired by using the chaos theory,and the tightness among the normal data and tightness measurement among the abnormal data are defined.Moreover,the tightness degree magnitude of data point in current each cluster is acquired,and the network environment containing large-scale redundant data disturb is suppressed by using the tightness and selectivity of data feature clustering method.Finally,the accurate detection of network abnormal data is completed.The simulation results show that the method has high detection precision.It can ensure the operation of network security and stability effectively.