CAJ | 학술논문

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。
침대당전수거분구알굴기술지고필수거본신고정특정이홀략료이상수거혼돈성특정이급계산이상수거특정성본교고적현상，설계료대수거혼돈성특정제취방안급취류분구알굴산법，실현료대수거적간화분구화이상수거적준학알굴。통과방진실험，험증료소제방법대불동대수거집가진행준학적이상수거알굴，병구유교고적로봉성화효솔。
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 .