前兆观测异常数据检测方法研究
详细信息 本馆镜像全文    |  推荐本文 | | 获取馆网全文
摘要
借鉴数据挖掘中的算法,本文设计了一种前兆时序模式表示方法,利用该方法可以快速检测数据序列中大幅突跳、阶跃等比较明显的异常数据。实际观测数据应用结果表明,该方法对于大量数据的异常检测效率很高,对前兆数据的预处理工作具有积极意义。
Based on the data mining algorithms, a pattern representation method was designed to express the earthquake observation time series. By using this method, the obvious abnormal cases could be detected rapidly, such jumping sharply, and steps, etc..The results of the actual observation data application shows that this method works very effectively on the anomaly detection of massive data. It makes significant senses for the preprocessing work of earthquake precursor observation data.
引文
王建国,崔晓峰,陈化然等,2007.前兆观测数据监视及异常自动识别软件系统.地震研究,30(1):83—87.
    张兴国,王子影,李胜乐等,2011.地震前兆数据异常自动检测报警系统.地震地磁与研究,32(1):101—108.
    周大镯,李敏强,2008.基于序列重要点的时间序列分割.计算机工程,34(23):14—16.
    Breuning M.M.,Kriegel H.P.,Ng R.T.,et al.,2000.LOF:Identifying Density-based Local Outliers//Proceedingsof the ACM SIGMOD International Conference on Management of Data.Dallas:ACM Press,93—104.

版权所有:© 2023 中国地质图书馆 中国地质调查局地学文献中心