基于大数据混沌特性的分区域异常数据挖掘  

Regional Anomaly Data Mining Based on Chaotic Characteristics of Big Data

作　　者：郑羽洁[1] 李茜[1] 

机构地区：[1]广西经济管理干部学院计算机系,广西南宁530007

出　　处：《河池学院学报》2015年第5期68-73,共6页Journal of Hechi University

基　　金：广西教育科学"十二五"规划2015年度立项课题(2015C455)

摘　　要：针对当前数据分区挖掘技术只考虑数据本身固定特征而忽略了异常数据混沌性特征以及计算异常数据特征成本较高的现状,设计了大数据混沌性特征提取方案及聚类分区挖掘算法,实现了大数据的简化分区和异常数据的准确挖掘。通过仿真实验,验证了所提方法对不同大数据集可进行准确的异常数据挖掘,并具有较高的鲁棒性和效率。In view of the current situation that data partition mining technology only considers its own fixed characteristics, ignoring abnormal and that calculating the abnormal data＇ s characteristics is at high cost , feature extraction scheme of big data＇ s chaos characteristics, and clustering partitioning mining algorithms are designed to implement the simplified partition of large data and abnormal data accurate mining. Simulation results verify that accurate abnormal data mining for different large data sets can be conducted by the proposed method, and with high robustness and efficiency.

关 键 词：大数据 混沌特性 分区域挖掘 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]